Wikipedia:Replies to common objections

Some people have very strong reactions to Wikipedia. Some are nearly instantly hooked, and they love the idea; others think the idea is so absurd as not to require any serious consideration. The present page is for the latter people (and this one is for the former).

There are, in our experience, a number of common criticisms of the Wikipedia project; we offer replies to concerns in the following categories.

1. Letting arbitrary Internet users edit my golden prose is absurd: Wikipedia will attract vandals, cranks, partisans, and other people who will ruin it
2. Many Wikipedia articles are of poor quality, and there isn't a peer-review process; no self-respecting intellectual would be associated with it
3. Good quality requires peer review and expertise, which Wikipedia doesn't have
4. Wikipedia's extrapolation to continued growth is dubious
5. Miscellaneous concerns

By the way, many of the criticisms levelled at Wikipedia are not unique to it, but are due to the fact that Wikipedia is, at bottom, a wiki. Many of the same objections have been made to other wikis.

1. Letting arbitrary Internet users edit my golden prose is absurd: Wikipedia will attract vandals, cranks, partisans, and other people who will ruin it

I can't imagine having my golden prose edited by any passer-by. It's mine--so why would I let others touch it?

We (on Wikipedia) don't individually try to "own" the additions we make to Wikipedia. We are working together on statements of what is known (what constitutes human knowledge) about various subjects. Each of us individually benefits from this arrangement. Unless we really are experts on a subject, or we want to do a huge amount of research (which, in fact, some of us seem to do!), we cannot easily singlehandedly state what is known on a given subject. But if we put our heads together, we manage--not instantly, but over time--to create surprisingly balanced treatments of various subjects. This has happened on numerous occasions on Wikipedia.

Now, one of the conditions of our working together in this way is that we not aim at stating merely our own views on any given subject, but that we see to it that all competing views on any given subject are stated fairly and sympathetically. Of course, extreme minority views on a subject tend to be underemphasized, e.g., in scientific articles. But since Wikipedia is not paper, i.e., because it's not limited in space, we can have as much information on crank theories as any crank should wish--on pages about those crank theories. They're just not presented as the truth, except as the truth about what the crank, absurdly or not, believes. We have a policy against writing from anything but a neutral point of view.

It should be left up to each person's reason to decide what is the right view on any given question. Given that, it is of paramount importance that the various competing views be stated as clearly, completely, and sympathetically as possible, so that each can be reasonably judged on its own merits, rather than according to what one biased person says about the subject. And so it seems extremely valuable to me to have a complete, neutral statement of different views on all different controversies.

Having huge numbers of people working on the same articles with the jointly-held aim of writing from this above-described neutral point of view is surely one of the best ways to achieve this aim. That in fact has been our repeated experience on Wikipedia.

Consider the following explanation from a Wikipedian:

I consider myself a rational person, and I do welcome the editing of my articles by others (nor do I consider them "my articles" after a while anymore). Let me give you an example. I thought I understood Goedel's incompleteness theorem pretty well, and since the then existing article was short and incomplete, I decided to rewrite it. Since then, several people have chipped in, sometimes rewriting a paragraph, sometimes critizing an omission, sometimes deleting parts. I didn't agree with all changes, but with most of them. No material is ever lost since Wikipedia stores all previous versions of all articles. So I reverted a few changes back. Overall, the article is now much better than I could ever have written it alone.

If we assume that the world is full of reasonable people and that collectively they can arrive eventually at a reasonable conclusion, despite the worst efforts of a very few wreckers. It's called optimism.

But if anyone can edit any page, what is to stop a crank from replacing a perfectly good article with some complete nonsense? Cranks are posting ridiculous theories on the Internet all the time. Why won't they just come here, post their idiocy and start flame wars and ruin everything? Wikipedia is going to end up like Usenet--just a bunch of flame wars.

In the first nine months of Wikipedia's existence, and we have had relatively few "crank attacks" and comparatively little mere argument. "Crank attack" is actually a misleading description--it makes them sound harder to deal with than they really are. It's pretty easy to just delete some irremediably crankish dross.

You might wonder, though, why we haven't had more visitations from cranks. This is enlightening to consider. First of all, let's distinguish the various gradations of crank stuff. At one end is well-meaning, but ill-informed and amateurish work. In all honesty, Wikipedia has a fair bit of that--but new hands (particularly, experts on the subjects in question) arrive and go to work, and the amateurish work is usually straightened out. At the other end is complete gibberish, which is almost instantly deleted as soon as it appears on the Recent Changes page. In the middle is the more difficult stuff--unsupported but reasonable-sounding information, strongly biased articles, and articles with a mixture of information and crankishness.

Wikipedia has a very strong policy against biased articles, or articles that present idiosyncratic points of view as the general state of knowledge about a subject. However, this does not mean that idiosyncratic points of view are silenced or deleted; rather, they are contextualized. The more idiosyncratic an entry, the more likely it is to be modified. Because there is no ownership of the information on Wikipedia, an individual is compelled to contribute information that is convincingly true. Thus, cranks who cannot accept critical editing of their writing find they have no platform and leave; those who are willing to present their interests in less-biased ways stop being cranks.

As for becoming a forum for argument, this is a bit more of a problem, but it is dealt with fairly handily by the social mores of Wikipedia, aka Wikipetiquette. Arguments on article pages get moved either to a corresponding talk page (e.g. talk:theory of relativity) or to a new article page which presents the arguments within a neutral context (e.g., operating system advocacy).

The argument on the talk pages tends to be centered on how to improve the article, rather than on the merits of various competing views. We have an informal but widely-respected policy against using /Talk pages for partisan wrangling that has nothing to do with improving articles.

Usenet lacks at least two features that are absolutely essential to Wikipedia's success: (1) on Usenet, you can't edit other people's work, while we can here on Wikipedia, thereby encouraging creative and collegial collaboration; or more strongly, on Wikipedia, there's no such thing as "other people's work", because there's no ownership of information; (2) Usenet does not have the possibility of peer pressure and community-agreed and -enforced standards, which Wikipedia does have. Moreover, Usenet is a debate forum. Wikipedia is, very self-consciously, an encyclopedia project!

So, surprisingly or not, in spite of the cranks, a lot of work gets done. Reams of encyclopedic material is written every day. We need more experts of all sorts on board, though, of course.

Still, any crank could write for Wikipedia. So I don't see what use Wikipedia is as a reference. There are many authoritative-sounding websites that say the Holocaust never happened, or the moon landing was staged in a movie studio, or even that the Masons really control the government.

It's a rather poor analogy to compare Wikipedia to such websites. You cannot correct those websites, no matter how wrong you know them to be, because they are written by people who would never allow their work to be edited without their permission. A lot of those cranks are impervious to argument or are simply uncontactable. They simply do not thrive on Wikipedia; we have observed that they are simply not in their element here.

So, theoretically, crank stuff can be added to Wikipedia, yes; but since it's immediately wikipediated, it's incorrect to infer that Wikipedia can't be used as a reference. (There might be other reasons for thinking Wikipedia can't be used as a reference, but this one isn't any good.)

Still, some cranks are very persistent. Suppose some guy writes up a crankish page on the Holocaust, or whatever, and they keep changing it back to their most recent saved version, so if I go there and try to look it up it will be a matter of who changed it last?

Generally, partisans of all sorts are kept under the gun. Wikipedians feel pretty strongly about enforcing our nonbias policy. We've managed to work our way to rough consensus on a number of different topics. People who stubbornly insist on an article's reflecting their personal biases are rare, and then they generally receive a drubbing.

If opprobrium isn't enough to stop people--and in the first nine months of Wikipedia's existence, that has been enough--we can always block the offending IP address(es).

Well, maybe you don't have very much to worry about from cranks per se, but there are plenty of ignorant people who think they know stuff and leave enormous gaps in coverage, and plenty of partisans who are all too eager to leave out information that is important to presenting a balanced view. They'll be all too eager to post to Wikipedia, and that's going to create huge gaps in your coverage, which will ruin the project.

Not infrequently the initial author omits crucial information, whether due to ignorance or malice. In many cases, again, this is fixed quickly--again, you must bear in mind that there are scores of critical people looking over the Recent Changes page every day. As a result, Wikipedia has fairly decent, balanced articles (like all of our articles, still in progress!) about abortion, Scientology, prostitution, and capitalism.

Another point to bear in mind is that Wikipedia is a work-in-progress, a draft, an "alpha release" if you will. It does have many important gaps. In fact, many Wikipedians believe that's a good thing: see AlwaysLeaveSomethingUndoneDebate. But it's constantly improving, and it is unreasonable to expect that something that began life in January 2001 will have every aspect of every important subject covered by now. This lack of coverage isn't due to ignorance, partisans, cranks, or anything else malicious--it's due simply to the finite amount of time that a finite number of people have been working on it.

OK then, maybe the cranks, partisans, and nonexperts won't mess things up too badly. But that's just now. You say Wikipedia is growing like gangbusters. Suppose it does. Then you'll start to attract the attention of more malicious elements. You won't be able to rely on the efforts of the (relatively few) contributors when that happens. All the noise will eventually be larger than any group of editors can handle.

We've been Slashdotted before and had huge amounts of traffic, and while there were a few "malicious elements," they soon find out that it's just not worth their while. There doesn't seem to be any good reason to think that this basic feature of Wikipedia is going to change as it grows, whereas there is evidence to support the belief that it won't change.

The more people there are to abuse it, the more people there are to ward off the abuse: if all the writers are also editors (see Wikipedia policy), the few malicious elements are hopelessly outgunned. As traffic increases, so does the number of people who work on and care about the project. We do not have a static number of "editors" who are responsible for editing everything; the number of people who do editing-type work increases directly in proportion to the number of people working on the encyclopedia.

Moreover--and this is something that you might not be able to understand very well if you haven't actually experienced it--there is a fair bit of (mostly friendly) peer pressure, and community standards are constantly being reinforced. The cranks and partisans, etc., are not simply outgunned. They also receive considerable opprobrium if they abuse the system.

It's easy to underestimate the effect of separating the concept of information and individual ownership of that information; it is a profound change from the prevailing paradigm since the advent of the printing press and capitalism. In short, people act differently in cooperative environments than they do in competitive environments.

So there's no good reason, that I can see, to think that such standards cannot scale to a larger project; if they can't, though, we'll evidently have to deal with that problem when it does crop up.

Fine, so you have Recent Changes and backups to correct the occasional piece of vandalism, etc.--but what do you do if people start running scripts to repost their own bit of vandalism or spam, and from different locations so you can't just block their IP address?

This hasn't happened yet, and to the best of our knowledge hasn't yet happened on any wiki. Anyway, you need to be a wiki administrator (with a special password) to permanently and totally delete pages. Any user can remove the text of a page, but any other user could restore the text from the "kept pages" archive (which is kept for at least two weeks in the default wiki setup). If someone did an extensive attack, all offending IP addresses could be blocked from further editing by the admins. In the future a new feature may also allow an admin to undo all recent edits from a particular user or IP address.

If necessary, we can always restore yesterday's version from a backup we make of the server itself. This really isn't a problem; none of the sysops are worried about it.

However, this can become a somewhat serious issue; no one wants an escalating arms-race of the sort that can be found between the Slashdot admins and trolls, where both sides use ever more-complicated filters and scripts; or destruction like that has come to the IRC network, defenseless against true maliciousness.

OK, even if you have some effective (if nontraditional) quality control mechanisms in place, there's a special problem Wikipedia faces. Most of the contributors to this thing are "geeks" of various descriptions: hackers, scientists/academics, and so on. Such people tend to have tastes along certain lines. Similarly, in music you'd probably have good entries for J S Bach, or Dave Brubeck, or the Grateful Dead, but not for Johann Strauss, Glenn Miller or Michael Jackson. So how do you achieve "balance" when you have pages and pages devoted to the Dead or to Bob Dylan, and hardly anything about, say, George Gershwin?

This seems to be a perfectly legitimate concern. Basically, most of the people writing for Wikipedia right now are techie types, and others who for whatever reason are heavy Internet users. Among those people are (apparently) not many people who like to write bunches about Jane Austen (literature), Michelangelo (visual arts), gardening, architecture, dance, and theater. (Those are some of the weaker areas of Wikipedia right now. Come help us expand them!)

Our hope is that, when Wikipedia really hits the big time, while the percentages of people working on unpopular topics might remain the same, the sheer numbers of those people will be higher than they are now. The idea is that we'll be getting more content in those areas then. Besides, it's not as though we have a time limit. Even if the computer and mathematics areas fill up faster than the dance and literature areas, it doesn't follow that the latter areas will always be weak.

Another thing that we can do is target the weak areas and try to get contributors for those areas in various ways.

---

2. Many Wikipedia articles are of poor quality, and there isn't a peer-review process; no self-respecting intellectual would be associated with it

It seems like there should be a giant "under construction" sign on every page of the website. It seems worthless as a reference. I don't see what the point is.

Wikipedia is both a product and a process. As a product, right now, it may not seem all that exciting or even respectable. As a process, however, it is quite remarkable. Seeing it as a process, Wikipedia can be judged not by its state at any given moment but by how well it is growing, how well it is becoming what it will become.

One way of understanding the process is by imagining a perfect article, one that ranks as a 100 on a scale of 100. An article that does not yet exist would be a 0, and a stub article would be perhaps a 1. The Wikipedia process works by constantly improving the quality of any given article, such that any significant edit moves the article 10% closer to perfection.

A first edit might move an article from 0 to 10 on a scale of 100. Viewed as a product, an article with a 10 score must seem pretty lame. But as the process continues, the article constantly improves. The next edit moves it 10% closer, to 19, the next to 27.1, etc. As further edits accumulate, the quality of the article moves asymptotically towards perfection, and likewise the quality of the encyclopedia as a whole.

The people at work here think the Wikipedia process is remarkable, and they believe it's going to result in a fantastic product.

Surely it's not possible that very many upstanding intellectuals will want to participate in Wikipedia. After all, wiki software must be the most promiscuous form of publishing there is--Wikipedia will take anything from anybody!

But it is possible, because plenty of upstanding intellectuals do participate in Wikipedia. It's fun, first of all. But it can be fun for intellectually serious people only if we know that we're creating something of quality. And how do we know that? The basic outlines of the answer ought to be fairly obvious to anyone who has read Eric S. Raymond's famous essay on the open source movement, "The Cathedral and the Bazaar." Remember, if we can edit any page, then we can edit each other's work. Given enough eyeballs, all errors are shallow. We catch each other's mistakes and enjoy correcting them.

So, we're constantly monitoring the Recent Changes page. When a crank shows up and vandalizes a page, it's fixed nearly instantly. (We save back copies of all pages, and these are very easily accessible.) We (that is, we participants) work on a lot of different pages, and many of us feel some collective responsibility for how the whole thing looks. We're constantly cleaning up after each other and new people.

In the process, a camaraderie--a politeness and congeniality not found on many online discussion forums--has developed. We've got to respect each other, because we are each other's editors, and we all have more or less the same goal: to create a huge, high-quality free encyclopedia.

That's nice, but why should highly-qualified people get involved with Wikipedia? It's not peer-reviewed. So, isn't it lightweight? Why should any serious researcher care about it? Why should anyone rely on it? Wikipedia has a nice community, but it doesn't have much breadth, depth, or reliability; so if you want serious information, go to Britannica.

If Wikipedians believed that, we'd bag the whole thing. We think we are--gradually, and sometimes from very rough first drafts--developing a reliable resource. So what answer can we offer to the above concerns?

Part of the answer has already been given: Wikipedia's self-correction process (Wikipedia co-founder Jimmy Wales calls it "self-healing") is very robust. There is considerable value created by this open and public review process that is continually ongoing on Wikipedia--value that is very easy to underestimate, for those who have not experienced it adequately.

Another part of the answer is that, of course, we've been around since just last January, 2001. (Britannica's had a few centuries' head start.) Significantly, Wikipedia's rate of growth has been steadily increasing--in terms of article numbers and quality, traffic to the website, and attracting more highly-qualified contributors. So it seems very reasonable to think that within a few years the project will surpass Britannica in both breadth and depth. At our current rate of growth, we will have over 100,000 articles by 2005; articles begun this year will be, in all likelihood, fleshed out to great detail. Not a few articles already have been.

Another major answer is that Wikipedia is providing free, unlimited server space and well-designed page construction tools for anyone who needs to do something that fits within the Wikipedia mission and doesn't care about owning the information; a description that matches the prototypical academic researcher.

But what about reliability? That's a third part of the answer. It seems very likely that, sometime soon, Wikipedia will set up some sort of approval process, whereby certain versions of articles receive the stamp of approval of some body of Wikipedia reviewers. There have been two main proposals about how to set up a review process. Whatever the shape of the process, it would act entirely independently of article generation. (We certainly do not want to kill the goose that lays the golden eggs.) But after it's in place, we will be able to present a set of genuine expert-approved articles that can favorably compare with articles from any general encyclopedia--Britannica included.

Admittedly, Wikipedia is rather far from being the reliable resource that Britannica is. But it's growing beyond anyone's expectations. The rate of growth continues to increase. Once an approval process is installed, in short order Wikipedia will--many of us think--be able to boast a breadth, depth, and reliability to compare to any general encyclopedia you please.

Then we'll try to get to the depth and reliability of a whole reference library full of specialized encyclopedia--something no general encyclopedia has ever done.

Grand, but I looked at an area that I know something about, and I found all sorts of errors and omissions. I was surprised and amused. I obviously don't want to be associated with something of this low quality.

We certainly do not hide the fact that a lot of articles need a lot of work. We started just last January; the Wikipedia article creation process has been very robust and effective over the period of many months.

We too deplore bad work. But then we often just go ahead and fix the problems we see. It would be great if you would help us increase the quality of the project by doing the same. Yes, there is a lot of mediocre stuff that has been added. But Wikipedia is not finished; think of it as one giant rough draft of an encyclopedia, written by a bunch of different people. And despite all that, much of what you'll find here, if you explore around a bit, isn't half-bad considering its youth, and some of it is quite good.

Initially, some areas of Wikipedia might indeed look like a deep morass, and article by article, bit by bit, you try to establish dry ground, but it seems too overwhelming. It is fun though; if initially you don't want to be associated with it, you can always choose an anonymous handle. The whole concept of authorship is not germane to wikis anyway. Bad articles cannot be credited to you because Wikipedia articles aren't credited to anyone!

We certainly don't expect everyone to want to jump on board. But if the main thing that's stopping you at this point is that some articles in one area of Wikipedia are of substandard quality, we'd ask you to come back next year, or the year after. By that time, it seems pretty likely that the mistakes in those articles will be corrected, and a lot more details will have been supplied. In short, time alone will--it's reasonable to think--render the project something with which you'll want to be associated.

Maybe it will improve, and maybe it won't, but currently it's pretty lame. I looked up a topic I know something about and found just a few words, just a stub. That's ridiculous!

There are indeed a lot of "stub" entries (you seem to have found a particularly egregious example), and we share your opinion of their ridiculousness. But it doesn't make a lot of sense to judge the entire project, either as a product or as a process, on the basis of the lameness of the stubs. There are a lot of excellent articles on Wikipedia, too. The majority, at present, are just OK--not spectacular by anyone's measure. But that's all changing because we're constantly working on them. People are expanding articles all the time.

It's reasonable to think there will come a time when we have exhausted most of the most common subjects, and we've got at least "stubs" about almost everything. Then we have no choice but to get into things more deeply. We can look forward to that. All in good time!

It seems Britannica has extremely high standards for what they put into their publications, both online and offline. Wikipedia has no such standards. It's bound to be of shoddy quality.

It's simply false to say that Wikipedia has no standards--the standards we follow are those followed by each of its contributors, and in some cases, these are very high standards indeed. As we gain more traffic, we will continue to gain more expert help, and as gaps are filled in, the only way remaining for Wikipedia to improve in will be in quality and depth. This, in turn, is likely to attract more experts, who follow their own very high standards.

To make a claim about what standards Wikipedia follows is to make a claim about what present and future Wikipedia contributors follow; to say that such people have no standards is little more than a baseless insult based in ignorance.

---

3. Good quality requires peer review and expertise, which Wikipedia doesn't have

Brittanica is good not only because it is big. If that were the case, there would be no reason not to be satisfied with World Book or something of that sort. When it is good, Brittanica is so partly because it is authoritative, and it got that way by being selective. Wikipedia isn't selective; hence it will never be authoritative.

It's perfectly correct to say that Britannica is good not only because it's big; the high quality of its articles is very important. Certainly it got that way by having high standards. We can concede that, but what reason is there to believe that it is only "by being selective" (presumably by choosing who is going to write about what) is the only way to support and achieve high standards? Maybe there's another, more open way. Wikipedia is a good test of that proposition. We have, after all, managed to produce some really excellent articles--and, by the way, not all of these were written by the many Ph.D.'s and other highly credentialed people that we have working on this project.

Your experiment will probably not go well. Good quality requires peer review and expertise. Why should we care about the products of an arbitrary group of people whose knowledge and ability could range from expertise to hopeless ignorance? Ignorance mixed with knowledge does not benefit knowledge.

First of all, the hypothesis that openness is to the benefit of quality has already been tested, and to the benefit of the hypothesis: articles that have been worked on by many different people in the context of Wikipedia are now comparable to articles that can be found in some excellent encyclopedias. If, however, you insist on considering the hypothesis a priori, we hope you will ask yourself: which is more likely to be correct?

1. 1. A widely circulated article, subject to scrutiny, correction, and potentially constant improvement over a period of months or years, by vast numbers of experts and enthusiasts.
   2. An article written by a nonspecialist professional writer or a fair-to-middling scholar (as so many encyclopedia articles are), and not subject to public review and improvement.

Look, all this speculation and "experimentation" is fine and well, but if there's one thing I've learned in my studies, it's that you can't really evaluate the validity of a piece of nonfiction writing unless you know something about the author and his/her qualifications to speak on the topic--or at least you are provided with the appropriate references to support his/her claims.

That certainly does seem to be a reasonable thing to say, but there are a few different points to bear in mind. First, an increasing number of Wikipedia articles do have references, and this is something we broadly encourage.

Second, the greater the number of participants, the greater the sheer numbers of experts are involved in whipping our weaker articles into shape--so, while you might not know which experts have been at work on an article, if you know that an article has been around for many months and that we have some experts in the general area at work here, it's fairly likely that it's been given a going-over by those experts. In other words, knowledge of the process, and of the fact that it includes participants who are expert in a wide variety of subjects, is potentially a substitute for knowledge that some particular (alleged) expert has written some particular article. Perhaps the relevant question to ask is, "How expert is the community of people who have created Wikipedia?" The answer is, "We've got experts in several different fields, and new highly-qualified people are arriving all the time." We don't require that most or even very many of experts on this and that join us, or think well of us; we require only a few, who have been steadily "raising the bar" from the beginning of the project.

Third, we are at work on a new version of our WikiWiki software, and when it is ready, we will be able to install an approval process. The experts among us--and other experts who are willing to act as reviewers for the project--will be able to pick out particular versions of Wikipedia articles that are especially authoritative and trustworthy. For more information, please see Wikipedia approval mechanism. Now, the mere fact that such a mechanism is possible puts the lie to the claim that "you can't really evaluate the validity of a piece of nonfiction writing unless you know something about the author."

Indeed, then, I should like to see some means of peer review before edits are accepted on articles which have already been approved by some similar process of peer review. At the moment it is entirely in the hands of an individual whether he thinks a modification he intends is an improvement, so there comes a point when a modification is as likely to damage the resource. If some system could be installed, then you would protect against crank attacks as well as misjudgement, and ensure a continually improving resource.

As a community, almost all of us are opposed to what has been called the policy of completely "freezing" particular pages--so that they can be edited only by a select group of people (e.g., only the author and an "editor"). We feel that our own collective monitoring of Recent Changes is an adequate safeguards against cranks--see above. Moreover, it is quite obvious that Wikipedia has achieved what success it has so far precisely by being as open as it has been. So--again--we don't want to kill the goose that lays the golden eggs.

That said, perhaps someone who has the above suggestion will be pleased by the approval system mentioned above and which can be found discussed at Wikipedia approval mechanism. Such a system would identify a body of experts that would put its official stamp of approval on some articles. Those articles could still be just as easily revised as they were before, but there would also be a version that would be presented as the "approved" version. This way we can "freeze" high-quality content without freezing the process.

---

4. Wikipedia's extrapolation to continued growth is dubious

Many of your replies seem to assume that quality will improve as the website grows. But quantity doesn't always beget quality. There is simply no reason to suppose that more articles is automatically better.

Actually, there is, at least in Wikipedia's case. There are at least three reasons to think that increasing numbers of articles and participants will lead to higher quality.

First, the more people are participating, the sooner we hammer out basically-OK articles on all the easy topics, thereby making the project of more interest to specialists who are turned off by the obvious omission of basic and reliable information on easy topics.

Second, the more eyes see our articles, the more transparent the errors will be (over the long haul). While we might have one or two philosophers on board during one month, a year later we might have ten or twenty--and then mistakes in their work will be caught much more quickly.

Third, statistically, the more people are participating, the greater the sheer numbers of experts; that seems to be our experience so far. Moreover, as a matter of fact, people usually tend not to touch articles they know nothing about, particularly when the article is well-developed or when they know that some resident expert will pounce on their mistakes. (There are exceptions, of course.) So, the greater the number of participating experts, the higher the overall quality of the content produced under their general guidance. It is not mere hype to say that Wikipedia caters to the highest common denominator--it's actually an observation we've made!

It seems pretty foolish to make a simple extrapolation from past growth to future growth. It's easy to grow at a 20% growth rate for a few months, or even for a few years--but, of course, not indefinitely. More generally, it's surely fallacious to suppose that the growth rate in the past is any very good indication of what will happen in the future.

It seems our critic here believes we have the following simpleminded argument: "The number of Wikipedia articles has been growing at rate R for the past nine months; therefore, it will continue to grow at the rate of R for the indefinite future." If that's all there were to it, that indeed would be foolish to say; but that's not all there is to it.

To be clear, we agree that it's very risky to make any specific predictions about growth rates. But it does seem reasonable to suppose Wikipedia will continue to grow at a rapid rate.

Now, what makes it reasonable to think that Wikipedia will continue growing at a rapid clip is not simple extrapolation, but observation of the factors that have made it grow at the rapid clip so far. Google has been sending us lots of traffic (thousands of visitors a day from Google alone; it used to be just in the hundreds). The more traffic Google sends us, the more people get on board and create content; and then Google sends us even more traffic. Moreover, more and more people are linking to Wikipedia. This raises Wikipedia's Google rankings. (Thus more traffic, thus more content.) Already, plenty of Wikipedia pages are listed on the first few pages of Google results.

Now, that's only part of the argument. The other part is that, while there is attrition (some old contributors don't write so much anymore), there's an overall increase in active population. There's a lot more active Wikipedians now that there was, say, three months ago.

Another part of the argument is that the overall quality of Wikipedia has been increasing, and our experience so far indicates that it will, probably, continue to increase. This makes it more likely that people will take notice of the project, link to it, use its contents (properly sourcing Wikipedia), etc.

In short, "the rich get richer." Please note, this is not mere speculation: it's an explanation of how Wikipedia's growth has occurred in the last nine months.

Of course, we will run out of topics sooner or later--the number of encyclopedia topics is not infinite. But it is really huge. A lot bigger than 100,000, and a heck of a lot bigger than the number of topics contained in Britannica. Even if we reach a point at which we cannot grow significantly in breadth, we will still be able to grow significantly in depth.

---

5. Miscellaneous concerns

Why is there a need for an encyclopedia at all, in the age of the omniscient Internet? Why not just go to your favorite search engine and search for whatever topic on which you're looking for an encyclopedia article? You're more likely to find more information, including more interesting and more current information.

Here's a glib answer: isn't it interesting that, in fact, thousands of people per day arrive at Wikipedia via Google?

Here's a longer answer. The point made here cannot be denied: the Internet, armed with good search engines, functions not unlike a giant, and exceedingly useful, encyclopedia. But does it follow from that that there is no need for an open content, community-built encyclopedia? No.

There is such a thing as an intellectual division of labor--not only among people, but among types of books and among types of reference materials. One does not typically consult a dictionary when one is looking for an encyclopedia article, even though, sometimes, the required information is included in the encyclopedia. Similarly, sometimes an encyclopedia article is precisely what is required. Encyclopedia articles can be generally expected to include certain kinds of information, and only that information. Among other things, they are generally expected to be written from a neutral point of view.

Moreover, Wikipedia is free and open content--which this is valuable. This means that anyone will be able to use the content for any purpose, particularly for educational purposes. Given the reliability that an approval process would ensure, the prospects of the use of a really huge, free encyclopedia for educational purposes--including the development of specialized educational materials--is very exciting.

Finally, it is possible that in the fullness of time Wikipedia will contain more relevant, reliable information on any given topic than can be easily found via a search engine search. That's certainly our plan for it.

WikiWiki syntax is inadequate to the task of writing an encyclopedia. The trouble with all wikis is that they are all limited by their markup language. Encyclopedia need a more flexible, visual mechanism than just an ascii pseudo-HTML. There is an upper bound on how useful a wiki can be without the ability to diagram, annotate, upload images etc.

First of all, on Wikipedia, normal tags such as <b></b>, <table></table> and so forth all function properly. In addition, images can be uploaded, although it does take human interaction (at present, they must be sent to jasonr@bomis.com--we're working on a file uploader, though).

The largest concerns have been in the mathematical section of the site, which indeed is limited, which at present is limited by HTML itself, which was never meant to display mathematical formulas. People are invited to work actively on the problems; for an update, join Wikipedia-L and post a question.

There are a lot of feature requests that would improve the wiki software we're running. Currently, a Wikipedia member (Magnus Manske) is rewriting UseModWiki so that it's custom-tailored for Wikipedia. The PHP Wiki is being tested at Meta-Wikipedia All of this software is free and open source.

---

Replies to Critics Talk

See also : Wikipedia