Thursday, May 03, 2007

The Wikipedia debate

By now, everyone would have used the free online encyclopedia that anyone can edit, Wikipedia. It's is a portmanteau of wiki and encyclopedia. As you might have already known, "Wiki" comes from a Hawaiian word for quick, and an encyclopedia is where you get "reliable" information. By using Wikipedia, it is claimed that one can find information on practically anything. Also notice that a google search on any topic tends to throw up a Wikipedia page in it's first three links.

What does this Google-Wikipedia synergy actually lead to? It means that, whenever you need to find information about something, and you Google for it1, there's a high probably that you end up on a Wikipedia page. So, the next logical question is, how does it matter? If Wikipedia is an encyclopedia, then it should be a good place to find information, shouldn't it?

This is where some people think there's a problem. By letting anyone edit, there's a fundamental flaw in Wikipedia, and it's the problem of credibility. This flaw is especially true in science related articles. The fact that anyone can edit Wikipedia, means that someone who's not an expert can easily edit and publish the article, without any peer review or scrutiny from experts. The proponents of Wikipedia argue, based on faith, that the good and accurate articles will end up surviving, and the erroneous information will be removed.

So, although Wikipedia has an impressive collection of fairly accurate articles, such Pink Floyd or Topsy, is it something you should use when you want to learn something for the first time? A number of critics, including myself, believe that Wikipedia is not the place to go if you need scientific information, because it cannot guarantee reliable, accurate or valid informaton.

From John Baez's2 website on physics 3


At any given time, the Wikipedia unquestionably boasts some very impressive articles on scientific topics. Nonetheless, its utility as an encyclopedia is fatally compromised by the existence of a surprising number of articles which appear at first glance to be sober, factual, and supported by impressive citations, but which are nonetheless a farrago of misinformation. The problem here is that only an expert may be able to spot subtle misinformation in articles on highly technical scientific topics. The point is that an "encyclopedia" which can be safely consulted only by those who are already experts on the topic at hand is not a true encyclopedia at all!

A steadily growing number of Wikipedia articles uncritically promote fringe science "theories" or scientifically suspect investment schemes, including a surprising number of devices which, if they functioned as advertised, would constitute perpetual motion machines. It is an article of wikifaith that Wikipedia articles will improve monotonically as more people contribute edits, eventually approaching a stable state of near-perfection. In particular, adherents of the faith allege that "army of watchful volunteers" will quickly spot and correct any misinformation. However, I have been tracking problem articles for almost two years, and I know that this faith is misplaced. Many of the worst physics articles have very obviously never once been edited by a genuine physicist. Even worse, I have carefully examined the edit history of hundreds of bad science-related articles in the Wikipedia, and in dozens of cases, I believe that the evidence suggests that articles which were allegedly written by a disinterested volunteer, were in fact written by someone having a direct financial stake in the claims described in the article. (This kind of deception is known as "wikishilling"; it is simply the latest variation on a technique which has long been employed by scam artists.) These observations appear to be consistent with a general trend toward increasingly sophisticated and insidious attempts by pseudonymous editors to manipulate information presented in the Wikipedia in order to pursue some hidden personal agenda.

The trouble is that Wikipedia has erected almost no barricades to guard against such abuses. Indeed, many feel that, while traditional printed encyclopedias are biased toward mainstream knowledge — and justifiably so, since mainstream knowledge is the most stable and reliable — the Wikipedia is if anything biased against mainstream scholarship.


So, how does one best use Wikipedia? The best way to use it, would be with extreme caution. You could use it as a starting point of your quest for information, and verify whatever information you get with other independent websites. However, In the case of learning science, you're probably better of sticking to your book. A textbook is a structured, credible, peer-reviewed and best source of information if you want to learn something new.

If you're interested in the debate, there are some excellent articles written by various authors you can find in the External Links section.

Notes & External Links

[1] - And let's face it, you're hardly going to search in a different way.
[2] - Yes, the irony was intentional...
[3] - Misinformation Concerning Cosmology and Relativity



Some external links
Why Wikipedia Must Jettison Its Anti-Elitism - Larry Sanger, co-founder, Wikipedia
The Faith-Based Encyclopedia
On "Digital Maoism: The Hazards of the New Online Collectivism" - Jaron Lanier.
Avoid Wikipedia, warns Wikipedia chief


9 comments:

Santhosh Suresh said...

Have you seen this? http://xkcd.com/c214.html

I've never come across something more seriously useful to the general public in the upbringing of the average awareness quotient.

Wikipedia does have some quality standards requirements to be satisfied before any article is published, right?
Wiki My Own Summer (Shove It)!! I was under the impression such anomalies were a rarity. I am starting think otherwise...

Siddharth said...

/*
Wikipedia does have some quality standards requirements to be satisfied before any article is published, right?
*/

The wikipedia quality standards (such as WP:V, WP:NPOV, etc) depend many times on the interpretation of the poster and aren't enforced in many situations.

The main issue here is that there's a tradeoff between convinence and accuracy. Sure, wikipedia is a nice , convineint resource to improve general awarness, but is it something you turn to if you want to learn new, scientific information? IMO, the answer is a firm no.

seanachai said...

This study by Nature a while ago seemed to show Wikipedia was only a little more error strewn than Britannica.
I remember one study that claimed Wikipedia was more accurate than Britannica but I couldn't find it.
As an article in the Economist discussing the findings pointed out, the fact that there are any errors at all in Britannica is shocking.

Siddharth said...

That Nature study is very controversial.

For example, see
>this article
which highlights some of the problems in that study.

Although, I was very astonished at the number of errors in Britannica.

Mohan K.V said...

I beg to disagree. Far, far more words have gone into _describing_ the problems that _could_ exist in a system like Wikipedia, than into finding out _where_ *exactly* such misinformation is, in other words, _evidence_. For example, in none of the links you have given, or the articles I have read so far, have I come across anybody giving a link to a wiki article (or static links to a particular revision) speaking bad science. I personally think most of the 'comparative studies' between Wikipedia and Britannica were pathetic, and of no concern whatsover to people interested in Science. I remember one of the most touted errors was a wrong birthdate of a guy born in some 1150 AD.

I do agree that some degree of care must be exercised when citing Wikipedia as a source; but, I think every article written on this subject misses out on a very important _reason_ people use wikipedia: to know things just for the joy of it! For example,tell me where _else_ I could get to know about such _delightful_ stuff as, say,
Graham's number?


I think that all exhortations of 'beware of wikipedia' are pointless: people who use it know enough on how much to trust it.
:-)

Siddharth said...

There are many articles in wikipedia with bad science.
For one example,
- Hydrino Theory

The point I'm trying to make is that, since anyone can edit wikipedia, it isn't the place you want to visit to learn facts. You could use it to get prelimenary information about a topic, or browse till you find something interesting and then use alternative sources to check the facts.

Mohan K.V said...

I still say `alternate sources' are no better. Oh well, the evidence is anecdotal for either case, so lets use them both :)

Mohan K.V said...

I came across this book recently, just have a look at the Amazon page and the book's webpage:

http://www.amazon.com/Final-Theory-Rethinking-Scientific-Legacy/dp/1581126018/ref=sr_1_1/002-8259604-0508000?ie=UTF8&s=books&qid=1184716843&sr=1-1

http://www.thefinaltheory.com/homepage.html

ROFLMAO :D

Siddharth said...

Ha, that's crazy! What's sad is the rating. 4 and a half on 5?? Even Resnick & Halliday got only 4.

Btw, I've moved my blog to http://parseval.wordpress.com