From Wikipedia, the free encyclopedia

"There are only two truly infinite things, the universe and stupidity. And I am unsure about the universe." ::-Albert Einstein

"Life's but a walking shadow, a poor player

That struts and frets his hour upon the stage

And then is heard no more: it is a tale

Told by an idiot, full of sound and fury,

Signifying nothing.'"

Macbeth in Act V, Scene V, The Tragedy of Macbeth

"What a piece of work is a man! how noble in reason!

how infinite in faculty! in form and moving how

express and admirable! in action how like an angel!

in apprehension how like a god! the beauty of the

world! the paragon of animals!"

Hamlet in Act II, Scene II, The Tragedy of Hamlet, Prince of Denmark

Contents 1 Is Wikipedia fatally flawed? 1.1 Scalability issues 1.2 Environmental issues 1.3 Financial issues 1.4 Equality issues 1.5 Differences in culture dramatically affecting output 1.6 Meta discussion disorganization 2 Other issues 2.1 Policies can be changed by anyone 2.2 No ArbCom oversight 2.3 Vandalism 2.3.1 Throttling frequency of edits by IPs and new users 2.3.2 Ability to block sockpuppets automatically 3 Some thoughts 4 Related links

[edit] Is Wikipedia fatally flawed?

In asking such a question, it is not to question the value and vision of such a project. It is more to consider the philosophical question of how such a project fits within the grand scheme of human existence, and whether such a project so formed can long withstand the might of its own creator. Throughout history there have been innumerable attempts to create repositories of human knowledge. From the Great Library of Alexandria forward, humans have endeavoured to establish a hall of such great capacity that it could encompass the entire knowledge of humanity. Is such a dream achievable?

There are a number of formidable challenges which face the Wikipedia project. What follows below is, at this point, not a fully formed list nor is it ranked in any particular order. It should not be read as a diatribe; in fact quite the contrary. In stating all of this, I am in no way saying that I think working on Wikipedia is futile.

[edit] Scalability issues

I believe there are a number of scalability issues which will force community functionality to its knees.

• Mediation: As has been noted by several users, the current mediation and dispute resolution process has become too bureaucratic to be effective. ArbCom is overloaded with cases, and cases proceed at an agonizingly slow pace. The problem will continue to get worse.
  • I've recently discovered Wikipedia:Dispute resolution reform. This is helpful.
• Copyright violations: Most admins do not pay attention to the possibility of copyright violations. The numbers of copyright violations continue to escalate, and this problem will only get worse over time. Recently, a new criteria for speedy deletion of untagged images was put in place to help stem the tide of image content that was not free of copyright concerns. There is a similar new criteria for speedy deleting textual content that is copyrighted. These are stop-gap measures only. Ultimately, copyright violations will so inundate Wikipedia that it that the Wikimedia foundation will eventually become the target of a costly copyright lawsuit.
• WP:RFA process: In the period spanning 17 weeks from June of 2005 to October 2005, the number of admins nominated per week steadily increased. A 4 week moving average had the average number of nominations at 7.25 per week after four weeks. The last average was 20 nominations per week. This is an increase of nearly 200% over the time period. At such rates, the number of nominations per week by Fall of 2006 will be 71 nominations per week. Two years from now, it will be 122 nominations per week. By 2010, 275 nominations per week. This assumes linear growth. However, linear growth patterns have not been the norm for Wikpidea. In fact, the growth patterns have been significantly faster than linear growth. Assuming that such large numbers of RfAs make the process unworkable, new procedures need to be put into place to handle significantly larger numbers of RfAs being put forth.
• Archives of talk pages in various namespaces have become insanely large. Again I refer to WP:RFA. Currently, there are 34 pages of archives under the talk page. This, after just two years. Ten years from now (assuming linear growth) there will be roughly a hundred pages of archives. What use are these? The onboard search engine does a very poor job of searching these archives. I find myself using Google to find lost entries on Wikipedia. This tool is largely ineffective in tracking results of prior discussions.

[edit] Environmental issues

The main Wikipedia servers are held at a location in Tampa, Florida. This location is susceptible to extreme environmental damage, similar to what was witnessed as a result of Hurricane Katrina in 2005. A number of decades long forecasts have predicted the next 20-30 years will see significant increases in the number and power of hurricanes developing in the Atlantic Basin. Furthermore, A study completed in 2005 at Purdue University regarding the environmental impact of global warming showed the gulf coast states will see significant increases in rainfall patterns. While insurance and off site maintenance will ensure no great loss of data to the Wikipedia projects, there will be ongoing significant interruptions and possible catastrophic damage to the Tampa hosting location of the Wikipedia project.

Anecdotal:As I write, Hurricane Wilma is sweeping across the state of Florida. Editing on Wikipedia today has :been greatly slowed. Connection? Unknown. I do not know the status of the Tampa facility at this writing.

I believe the Wikipedia project would be better served by removing to a location less susceptible to environmental catastrophe. Furthermore, if the Wikipedia backups are maintained at some site in Florida, it needs to be moved to an out of state facility considerably less susceptible to flooding or other environmental catastrophe.

[edit] Financial issues

A frequent complaint on Wikipedia is the slowness of response by the servers. An oft heard response is that as servers are put in place to speed response times, so does traffic resulting in a net minimal gain in performance, if any gain at all. Yet, there are a number of companies within the top 50 of most visited websites that have been able to respond to the demand and still provide rapid response to their surfers. People are increasingly becoming tired of the World Wide Wait. Fast access to data must be the norm rather than the exception for websites of the future to succeed. Wikipedia has failed and for the forseeable future will fail to provide the service necessary to keep it as a front line top website.

[edit] Equality issues

Under the current model, a brand new user is treated within wikipedia as being identical to a user with years of experience. In an ideal world, this might be an admirable view. It encourages the notion that all people are created equal, and the qualities which make them good members of a community are in place from their very first edit. The rest of their behavior is a matter of training. However, there are serious flaws within this construct:

• There is little in the way of organized efforts to train people. Currently, there is a Welcoming Committee, but it has no capability to ensure all new users are welcomed. Within the welcome messages that users place on new user's page, there is disparity as to content of those messages. It could be as simple as "Welcome to Wikipedia" through to several paragraphs of suggestions, things to read, etc. These welcome messages are read with varying degrees of understanding by these new users. The results of these efforts are predictable; haphazard understanding by new users of what it is they are doing in Wikipedia.
• While the IBM study showed that Wikipedia stumbles forward towards accuracy, the burden of verifiability is very large and increasing.
• Prior to the Internet explosion which began in the early 1990s, access to services on the Internet was, outside of government and military avenues, restricted mainly to people coming from institutions of higher education. While there were and always have been idiots and trolls, the numbers were relatively small. With the introduction of services such as America Online, Compuserve and Prodigy access to the Internet was opened to every person who had a computer, phone line, and willingness to spend a small amount of money per month to have access. In the view of many people, this caused a great dumbing-down of the Internet. Wikipedia is suffering from the same problem. Under the current construct, the best that Wikipedia can hope to be is the sum of its contributors. Yet, the contributors are unregulated. As a result, the average contributing Wikipedian has, unsurprisingly, average intelligence. Juxtaposing the goal of creating an encyclopedia and the body of editors being of average intelligence leads to the highlighting of a dichotomy; how do you create a learned work using the skills and talents of people who are, on the whole, of average intelligence? Worse, how do you do so with the presence of a large number of people who are less intelligent than average?
• Similarly, demographic distribution of editors could potentially skew the net ability of Wikipedia editors in one direction or another. Without accounting for these differences, then a number of outcomes not inline with the overall goal will pervade the resource.
• The presence of emotion driven contributors is anathema to Wikipedia. For example, conspiracy theorists are highly motivated individuals, incapable of being diverted by rational questioning and requests for basis in fact. Nevertheless, they can and will continue to 'contribute' to Wikipedia. These conspiracy theorists often radically disagree with each other on core, basic facts in order to prop up their pet theories. Individuals hawking one theory over another insist that theirs is the only one, right theory and everything else is false. This is illogical, but it does not stop them from editing Wikipedia. Conspiracy theorists also routinely insist that their pet theory is valid until proven false. Well, nobody's proved my theory that death ray using Zygorthians shot down TWA 800 either. On Wikipedia, we strive towards concensus to achieve articles that we find to be the best we can produce. Yet, that concensus includes people who are probably certifiably crazy, driven by emotions, and motivated to conduct a huge number of edits to promote their pet theories. Emotion driven editors are given the same equality as the would-be Platos and Socrates of Wikipedia. 10,000 monkeys can not write Shakespeare, much less Wikipedia. Yet, that is what we're trying to attempt.

[edit] Differences in culture dramatically affecting output

Currently, the dominant population demographic on the Internet is people who speak English. This is slowly changing as access to the Internet becomes more ubiquitous. As a repository of information the various Wikipedia projects can and will differ with each other. Which is right? Is it possible to say which one is right? I would say that in many cases it is not. For example, the unrest in the Middle East has already yielded a number of disagreements on Wikipedia. What looks like an act of terroristic aggression to one party is an act of patriotic valor on the part of another. You say tomayto, I say tomaato, let's call the whole thing off.

[edit] Meta discussion disorganization

• A number of contributors 'participate' in Wikipedia by using the IRC channels. This participation is unrecorded, unreviewable, and outside of the bounds of any constructs within Wikipedia under which to manage it. Yet, a number of activities conducted on Wikipedia are the result of discussions happening on IRC.
• Talk pages on articles make poor forums for discussion of various aspects of articles, especially contentious articles. There is no forced signature (why not?), which can make it difficult to discern who said what without going into the history of the talk page (arduous, at best). Indenting is used poorly to reflect changes in who is speaking. Sorting by date is impossible. Sorting by thread is impossible. These talk pages are frequently a disorganized mess that make it difficult to understand core issues affecting the page or bring concensus to bear.
• There is little in the way of structure in place for people to review past decisions and how those decisions were arrived at. Within the jurisprudence of Western nations, there are frequent references to precedent decisions and the merits of those decisions are often used to support later decisions. Here, there is no such system in place by which an honest mediator can look to for guidance. Mostly, they are left on their own to interpret the policies and guidelines as they read them; yet such things are subject to interpretation. Frequently, people cite past behavior as a basis for a decision. Yet, when taken to task to show the basis in policy or even evidence of the past behavior, few people can produce it. This undermines effective decision making and forces long, repeated debates over subjects that previous had been debated ad nauseum. Wikipedia:Arbitration policy/Precedents exists, but I feel this is insufficient.

[edit] Other issues

[edit] Policies can be changed by anyone

One of the core aspects that governs user behavior on Wikipedia is the policies. Yet, these policies can be edited by anyone. Sure, people are encouraged to discuss substantial changes on the respective policy's talk page, but as previously noted talk pages make poor forums. If policies really are policies, then who can change them should be restricted. Else, the policies can and do evolve far too quickly for people to keep up to date on. It's like having a nation's tax law change every few weeks rather than once or less per year. Did a person violate policy? Depends on when the violation was and what the current policy was at that moment in time. What if a policy changes while a person is involved in a dispute? Does the basis for the dispute disappear if the wording being noted in the dispute disappears? A number of people say it is more important that we follow the spirit of a policy rather than the precise words. This is probably in reaction to the fact that policies are constantly changing. The problem is this leaves an understanding of what the "spirit" of the policy is up to the judgement of an individual editor, rather than community consensus. This sets our users up for trouble as we have an ever changing mass of policy they are expected to follow and it is a moving target, a target they are left to interpret on their own.

[edit] No ArbCom oversight

ArbCom's decisions are inviolate. Once a decision comes down from ArbCom, everyone on Wikipedia must abide by the decision (if the decision affects them). ArbCom is notionally an elective position, though the original 12 person slate of members were all appointed. The reality is the majority of ArbCom members never serve their full terms. When this happens, replacement members are appointed rather than elected into the positions. The consequence of this is the majority of ArbCom is not made up of people in whom the community has vested their trust, though at some periods of time in the future it may be. If an ArbCom member becomes unpalatable to the community, there is no means by which a member can be removed. As a result, ArbCom is not accountable to the community either in terms of electing or removing people from ArbCom. As it stands, a member of ArbCom can act in any way they wish without taking an account of how others might respond to their actions.

In stating the above, I do not suggest we should allow recall elections at any time for an ArbCom member. As it is, they are charged with working on issues that are virtually by definition contentious. People will be upset with ArbCom members more frequently than not. I respect that it is a dirty, thankless job. That said, I still believe there should be a process to remove an ArbCom member who the community has lost faith in. Perhaps a petition of x number of editors with more than three months on Wikipedia to initiate a recall election, followed by an election.

Peculiarly, we've had at least one ArbCom member who was appointed to ArbCom, failed in an election to remain in ArbCom, and then re-appointed just months later. Without knowing the details, a person could readily conclude that there is a member of ArbCom in whom the community has less than acceptable faith.

[edit] Vandalism

Vandalism is an ever increasing problem for Wikipedia. The more influential it becomes, the more vandalism will occur. Yet, the only solution we seem to have to deal with vandalism are the noble efforts of users reverting vandalism and the efforts of administrators. This is wholly insufficient in my opinion. The problem of vandalism will continue to get worse. We are already overloaded in contending with the problem, and the situation is getting worse, not better. We need better tools to combat vandalism. We are here to write an encyclopedia. Yet, more and more precious time of our good editors is being spent undoing vandalism rather than creating articles or adding good material to existing articles. Thus, even if all vandalism is undone, there is still a very large negative impact on the project because of the time that must be diverted away from productive actions to reactive actions.

[edit] Throttling frequency of edits by IPs and new users

Vandalism most frequently comes from anonymous IPs and newly registered accounts. We allow these people to edit as fast as they like in vandalism. We leave it up to the watchful eyes of RC patrollers to pick up on the activity and take action to stop it. It would make the job of vandal fighters significantly easier if anon editors and new users were restricted to one edit every five minutes. A user who wants to make a large contribution to a given article can still do so. Once they have made a small number of edits, say 50, then the throttle could be eased back to two minutes or removed entirely. This would have a dramatic slowing effect on vandalism.

[edit] Ability to block sockpuppets automatically

I do not feel that checkuser rights should be liberally handed out. In fact, quite the opposite. However, there should be a technological solution to be able to provide a tool where if an admin blocks a given user, all suspected sock puppet accounts of that user, say with less than 100 edits, are automatically blocked as well, along with automatic messages placed on the talk pages of those sock puppet accounts. A vandal could get around this, but if the above mentioned throttle were also part of the picture, the vandal would be greatly slowed.

[edit] Some thoughts

I would argue at this point that Wikipedia does not serve well as an encyclopedia. Rather, it is a highly interesting sociological experiment. I imagine there will be PhD dissertation work on this experiment at some point in the future.

[edit] Related links

• User:Kate/WhyWikiPediaWorksNot