Slashdot Mirror
Ex-Britannica Editor Reviews Wikipedia
0-9a-f writes "Robert McHenry, one-time Editor in Chief of Encyclopædia Britannica, offers his thoughts on Wikipedia at Tech Central Station. While many Wikipedia zealots might discount his obvious bias outright, his broad argument is difficult to ignore. A million monkeys might eventually write Shakespeare, but how would they recognise it once they had?"
Robert McHenry asked "how would they recognize it once they had (Shakespeare)"
Simple. For each Shakespeare literature there would be another million monkeys reading and discussing the article. Thus you have a million writing monkeys and you would have maybe a million million reading monkeys; thus, the noise from the million million monkeys during discussion would drive the million monkeys.
foreach $monkeys(keys {%Shakespeare})
{
print "You\'ve got Shakespeare" if %shakespeare{$monkeys} = $It;
}
See the infinite monkey rule isn't good to apply as that rule doesn't facilitate feedback from the system.
I've been using Wikipedia almost exclusively as my encyclypedia for over a year now.
FTA:
The author says there are "no means to resolve" but I beg to differ. There is clearly a means to resolve these inconsistencies in that particular article! Edit it!! If he has found something wrong with the article, he should take a few minutes and correct it. Enough of that, and the article will go into dispute and moderators will resolve it. If this author is interested in Alexander Hamilton, he should watch that thread unfold using the Wikipedia tools to stay on top of it, making changes as he goes.
The nice part about a Wiki is that the changes are tracked, so the wiki on a whole is bigger than the page you are looking at. You can see how articles evolve, and where disputes may find fuel. Furthermore, this kind of thinking requires more depth than the printed page ever could.
When you are a dinosaur, you ought be extinct or you ought adapt, IMHO. Make way for the Humans! It's apparent to me that this author understands neither the concept nor the spirit of Wiki, and considering he is in the Encyclopedia business -- that is quite troubling, as it is mission critical for any field to understand new technologies as they unfold within that field.
The dangers of knowledge trigger emotional distress in human beings.
As an educator, Wikipedia needs to have impeccable credentials and support from leading educational institutions before I would recommend it to our teachers and students.
While many Wikipedia zealots might discount his obvious bias outright
Wikipedia is the most biased "reference" source out there. The Karl Rove ariticle basically made him out to be a reincarnated Goebbels. The problem of course is any editor with an agenda can ruin an article.
As for the facts, I've seen howlers in many mainstream encyclopedias. In the cases I know something about, I find wikipedia's standards quite good, and when there's an error I can at least go in there and correct it.
It's true I crosscheck anything I find there but I do that with other sources too. Never rely on a single source.
So true! Thats like saying a million monkeys might write a great open-source operating system, but how would they recognise it once they had?
ermm.. wait...
Wikipedia's process for moving from an idea to a collection of badly edited articles to a real encyclopedia is, at the risk of soundling like someone from the 90s, exactly the same as the process by which any community learns.
On an infinite timeline, Wikipedia is going to beat the snot out of anyone else--in about 200 years, it will have incorporated everything written before the 21st century into itself.
To speed it along on a realistic pace, the only things that can be done are either contributions or, *gasp*, donations specifically earmarked to hire fact-checkers and editors.
That it's implausible to suppose that a large community of contributors might eventually write an operating system that could challenge Windows in the market.
Of course, the comparison isn't completely accurate, since Linux and *BSD do have "gatekeepers", people like Linus and lieutenants, who at least in theory are vetting everything that makes it into the main kernel.
Nonetheless, it's not a million monkeys writing Wikipedia. Many are monkeys, but there are also lots of intelligent peope out there.
It's also naive to suppose that every "traditional" encyclopedia article has been completely free of error. (Just as naive as the assertion that Microsoft's quality control makes Windows free of security holes.)
Sure, Wikipedia isn't perfect. Sure, it's very easy to see how bad information can get in there (not even creep in, but stroll in through the front door and sit down). But if enough people are buying into it, it's also easy to see how the process can work. So far, by and large, it seems that it is working, even if not perfectly.
Given that (at least until various regulatory agencies and large intellectual property firms manage to codify their horror) the Internet allows everybody to be a "content producer", not just those who control the huge resources of a publishing company, it's only natural that there should be a sort of encyclopedia that allows each to contribute his own expertise without going through the priesthood of a encyclopedia editorial board. Will it make traditional encyclopedias obselete? Certiainly not, at least in the short term! But nor do the differences mean that something like Wikipedia shouldn't exist and that people searching for information should eschew it in favor of traditionally published encyclopedias.
The future (longer term) of encyclopedias will almost certainly look much more like Wikipedia than traditional encyclopedias. Perhaps they will have a "small" set of gatekeepers (a la Linux), but they are almost certainly going to be ready and willing to accept voluntary contributions and edits from all and sundry, just from the very raw point of view of efficiency and harnessing as diverse expertise as possible.
-Rob
"...the process allows Wikipedia to approach the truth asymptotically..."
This is perhaps the most compelling point made in the article, to me. Of course, the cynic's read into that statement is that Wikipedia will never get to the truth (see Asymptote). In some ways though, that's really a pretty undeniable truth about the Wikipedia system -- even if it is True today, some jackass can come in and make it Not True tomorrow. Even if it's Not True for only five minutes, if someone looks at it during that time and assumes it to be correct, the wiki has failed in some sense.
Don't get me wrong, I really love Wikipedia, but I think some of the points raised a very much deserving of further discussion -- if you can make a crofty old coot like this guy happy, it's probably going to be a pretty damn good [encylo|wiki]pedia.
If the "monkeys" decide they like what they wrote, that's good
But the problem is that more often than not the monkey's opinion of the truth or fact isn't in fact necessarily congruent with the truth or fact.
Popular isn't necessarily correct or incorrect. It's just popular. You can have a dozen wikipedians arguing back and forth on a topic but at the end of the day the socratic or arugmentative process doesn't guarantee a solid article.
The word "bias" gets tossed around a little too much in American discourse these days. How, pray tell, might we honestly construe this man as biased?
It isn't "biased" to be educated or to have the experience necessary to provide a thoughtful and determinative analysis.
Indeed, this man's entire lifetime has been dedicated to editing a series of books whose entire modus operendi is to present information factually and to be explicitly aware of their own limitations. An encyclopaedia is by defination a reference work, a limited collection of reliable information that leads you to further study. That is the opposite of "biased", which is to present self-serving conclusions based on a self-serving assemblage of information.
One thing many Western societies lack right now (but, I would offer, America in particular), is widely accepted basis for producing legitimate knowledge. There are serious concerns with the Wikipedia as a source of authoritative information that exacerbate this problem, not address it.
I welcome this man's comments rather than condemn them.
"If the 'monkeys' decide they like what they wrote, that's good enough -- it doesn't have to be Shakespeare."
Your sole standard is whether you "like" what's written?! It appears that truth no longer matters in your bottom-up society.
If someone says he and his monkey have nothing to hide, they almost certainly do.
This has nothing to do with elites and top down society. The point of the article is that Wikipedia may not be the ultimate encyclopedia as some of its boosters may proclaim.
To address your point, you're saying that tyranny of mediocrity is acceptable and in fact desireable? In your world, there is no reason for people to aspire to higher knowledge and enlightenment?
Self awareness - try it!
I don't really get why some people get so upset over WIkipedia, and wants to defend ordinary encyclopedias as "more authoritative".
When it really matters, Wikipedia is of course not a primary source to go to. But then, neither are ordinary encyclopedias. When it _really_ matters, you go to the original research papers, subject-specific anthologies and conference proceedings. You will likely never see Encyclopedia Britannica referred to as an authority for an FDA application, for example, or for an envrionmental consequence analysis for some proposed industrial development.
What encyclopedias are good for, on the other hand, is to give a quick tour of and route into an area the reader isn't already familiar with. And since any deeper delving into the subject will require referencing a lot of other sources in any case, any smaller biases or omissions in this "portal text" isn't going to matter.
Trust the Computer. The Computer is your friend.
And as far as inaccurate information goes, I have a two word response for that: political blogs. Many people are perfectly happy to get their Important Information a blog by somebody who can't name their sources and who has no responsibility to be accurate. The modern measure of accuracy is simply a matter of how many people believe and repeat a statement.
Err... you confuse accuracy and popularity.
The modern measure of perceived accuracy may be that, but that doesn't make it right.
If enough people believe that the world is flat, that won't make it so. Even if lots and lots of blogs talk about it.
-Rob
Indeed. If he had fixed it "instead of brewing up some fluff piece", "the world would be in a much better place" AND you wouldn't have entirely missed the point of his fluff piece.
I still remember the encyclopedia salesman that would set up in the mall. Heck, we even have a couple of very nice encyclopeidas in the house.
The problem is that information becomes dated very fast. Encyclopedias are useless for researching anything technology related, except as a historical snapshot. And with the collapse of the Soviet Union, new countries were springing into existance faster than the maps could be printed. Revolutions happen, presidents change and information that was once 100% correct becomes stale or downright wrong as new things are discovered. (How much more have we learned about Mars in the past year?) Despite the problems, online encyclopedias are still the way to go, and I would value Wikipedia as a reference far more than the beautiful leather bound dead tree editions.
My parents have a 1930's vintage encyclopedia set that they picked up in a garage sale once. It is quite facinating to go through and read a snapshot of what was known and believed to be true at the time.
My rights don't need management.
In other words, a hundred ill-informed opinions are still worse than one well-informed one. And simply having more people contributing to a piece of work does not necessarily make it better.
Looking for political forums? Check out "The World Forum".
I will say one thing Wiki excels at over traditional resources is Science and Technology. For example: The Oort cloud, which is a theoretical source of comets, is often gospel in many lower level science and encyclopedia text books.
Britannica Article
Wiki Article
As you can see there is a major difference in the way the theory is presented. Britannica as science fact and Wiki as theory.
Wikipedia is a community effort.
If we replace the word "community" with the word "committee" the problem is obvious.
http://rareformnewmedia.com/
But you don't get it. He has helped. He has identified weaknesses that few people have considered. He has brought his experience of editing encyclopedias in a commercial environment, where accuracy and adequate referencing is paramount, to the Wikipedia project - for free. The stupidest thing that could happen is if Wikipedians don't act on his comments and just whine "why didn't you fix the article".
OK. Imagine - he does what you ask. He fixes the article. The Wikipedia now has one fixed article and still has all the systemic problems it had ten minutes ago.
Now how in the hell am I supposed to trust this definition of Asymptote?
My beliefs do not require that you agree with them.
One of the most annoying things I find about Slashdot is the immediate reflexive response to regard an article as either 'for' or 'against' issue X. As soon as I saw that an old Brittanica writer had commented on Wikipedia I could guess the shape of the Slashdot debate, without even knowing what the Brittanica fellow had said.
I have read his comments, and as a not insignificant Wikipedia contributor, I have to say they're correct: he gets it. He does not regard Wikipedia as a useless adventure, but he does not trust (have ) that the collaborative process will necessarily produce excellent-quality articles.
I have to say I agree. I admire the idea that quality is a sought-after goal, but such efforts as the Collaboration of the Week succeed only because Wikipedians focus their attentions on a given article closely for a short period of time.
I have seen too many articles that are confusing and disorganized at a meta-level. A simple factual error invites itself to be corrected, and therefore will be corrected, but restructuring a whole article when you know someone may come along and violate your scheme tomorrow is a discouraging thing.
As well, too many articles on controversial subjects end up being a confusing mismash of argument against or for the point in question. This is particularly the case for recent controversial political figures. I'm happy all the information is there, but I will not believe that the collaborative process will naturally produce an article that covers the issue fairly.
I view the Wikipedia as analogous to a probabalistic algorithm in computer science (e.g. a probabalistic primality testing algorithm). Such an algorithm is true most of the time, and can be a hell of a lot faster than the always-true deterministic algorithm.
Those who criticize the algorithm's potential for falseness miss the fact that its nondeterminism gives it great power, but its proponents should never forget that it is not deterministic.
I am a research scientist and the material that can be found on wikipedia's website in the subjects of phyics and mathematics is vastly superior to anything the commercial encyclopedias have published. They seem to focus on creating material for high-school students, but their texts are largely useless for higher level physics and mathematics. They just don't have enough detail. This is where Wikipedia excels. Although Wikipedia's converage of physics and mathematics is often written in terms not familiar to a layman, there is often some part of the article that makes it understandable to those who are not involved in the fields of physics and mathematics.
Thumbs up to the guys at Wikipedia and to those who have contributed articles on mathematics and physics.
I also wouldn't trust it not to gloss over important aspects of topics and to create the impression that a relatively unimportant aspect of a topic was more important than it really was by going into too much detail over it.
I could say the same for Wikipedia. Except that I haven't cracked open an encyclopedia in years whereas I use Wikipedia three or four times a week to look up a fact. Most of the time I don't go directly to the site, but search for the topic using google, and then click on the link to a wikipedia article that will show up. I know the link is worth clicking if it comes from wikipedia or one of the advertising supported 'mirrors'. I don't even mind the ads since I mostly browse with lynx anyway.
But I wouldn't feel super confident that what I read in a wikipedia article was the complete and total truth ( though most of the time it comes close ) until I had at least checked out a few other sources.
Sometimes, I used to start at the 'top level' of a subject in wikipedia that I wanted to learn about, and then click the links, going into as much depth as I felt like by clicking ever-deeper. The text was structured as an article, and the subjects that were links were in context. I loved this because it made learning about a subject in general easy. Now that wikipedia seems to have reorganised it's top levels by deleting the well written and informative top level articles and replacing them with information-barren alphabetic indexes, that sort of learning is not as easy, though it can still be done once you go a little deeper into the articles.
In my opinion, the alphabetic indexes should have been added to, wikipedia, but not replaced the top level articles which put the subtopics so nicely in context.
It's a sifting of global consciousness on a certain level.
What does the average computer user think about, 'X'? You can get a pretty good idea with Wikkipedia. Then, because it's the internet and EVERYBODY should by now recognize that when doing research on the web, one needs to read a bunch of different websites on the same data they're exploring, research the owners of the website to see what their inborn bias is and what other things they have done, and then do a bunch of creative cross-referencing work. For some subjects, it provides and excellent starting point, but in the end, further research should always include more and wider explorations. The same must be said of any body of reference material, including Britannica.
And, of course, if you need the orthodox viewpoint written from Official Culture, spun to the tune of "Nothing to see here, citizen", then by all means, look up Britannica. (I particularly liked the difference between the two definitions for the word "Orthodox"; Note particularly, the first sentence on each; Wikki gives us an actual definition, whereas Britannica starts out by immediately telling us that Orthodox means, "True". The irony is downright chewable.)
"Orthodox"
"Chemtrail"
-FL
In brief, at the end of the day after 100+ edits, the Alexander Hamilton piece is NOT a rich tapestry of nuance and expertise. It's a high-school quality wallpaper job.
The author has proposed mechanism as to why such articles are, in effect, wallpaper jobs and does, in my opinion, a good bit to evidence the "emperor has no clothes" nature of those such as yourself who have a faith-based view of collaboration - the well meaning, but certainly not proven and possibly quite wrong idea that groups of humans "quasi Darwinially" converge upon optimal solutions.
The probem may not be that the author doesn't understand the spirit of Wiki - it may be that he understands it too well.
/ full disclosure: I have contributed articles to Wiki, though I am under no illusions as to its potential and, frankly, share the author's views. When I do serious work, I don't use Wiki as a reference.
Any article can be corrupted to any truth distance value at any step of the process. In addition, there is no guarantee that eventually corrections would be made. And if there are useful corrections, there is no guarantee that they too won't be undone.
It's as if Slashdot decided to use only the last moderator to determine whats insightful, interesting or funny.
A sequence of random numbers doesn't converge. Of course, an inifinite set of radom number sequences might contian one that does ...
True, but I would argue that using a single source, including Britannica, is just an incredibly unwise thing to do in the first place. If it's important enough to matter you would be a fool to use a single source. Even the oh so holy Britannica has it's biases and omissions.
But how good was the EB in its 5 year of publication? I bet they were publishing phrenology as a real science. ...I bet that pompous jerk didn't even take the 3 minutes to correct the Alexander Hamilton article.
EB has demonstrated a successful method for creating a great encyclopedia. (It is safe to say that after phrenology was debunked, EB did not continue promoting it; indeed, it is likely that EB was skeptical even at the time.) EB hires experts to write and uses experts to review. EB charges its readers money for their work.
We hope to demonstrate that a great encyclopedia can be created by open collaboration of uncredentialed volunteers. If we succeed, Wikipedia will be a powerful example of the applicability of open source methodology outside the software arena. To date, we have not succeeded. We all know that Wikipedia still falls woefully short on many of the key criteria for a great encyclopedia (accuracy, breadth, depty, currency, grammar, cost, bias, etc.). (The article's point about Alexander Hamilton is well taken. Not only did yesterday's Wikipedia contain errors on simple matters of record of Hamilton's career, but our collaborative editing process had made things worse, not better.)
If we want to prove our hypothesis (that the bazaar can create things just as well as the cathedral can), then we must not only keep contributing, but we must keep refining our editing protocols to prevent the kind of negative progress we saw with the Hamilton article. It will not help to call our opponents "pompous jerk[s]" for not doing our jobs for us. They've proven that their methodology works. Now, we must prove that ours does.
"We reject as false the choice between our safety and our ideals." --The American President (20.1.2009)
Wikipedia, when it is at its best, is similar. It will never reach the Truth, however, as people contribute to it, it will hopefully approach it. Information that is not useful (because it conflicts too grossly with other "models of the Truth" out there) will be removed, and information that is useful (help the users) will be added.
The Britannica author comes from another tradition. A tradition where Truth is based on authority rather than consensus. The ultimate Truth is God, and is expressed through the hierarchy of the Church down to the common churchgoer. Lately, the Church has been supplemented by Science. This gives the common layman view of Science as a Truth, competing or supplementing the Church. Scientists, of course, know that is not so, but the whole dissemination system (schools) has not been updated yet. It uses the old Church based mechanisms. When scientists teach, they try to teach pupils to think. They don't just pass knowledge given from above.
Much of the Britannica authors ruminations about the degeneration of modern society stems from the same source. Focus is shifting towards the process, and old barriers are removed. Teaching methods is (slowly) catching up. The world is changing, and the best you can teach your pupils is how to adapt to the change. He does not understand that. What was once the Truth, will always be the Truth. That is the nature of Truth. He complains that Wikipedia does not consider the reader, only the authors. This is because the Wikipedians don't use the same model of the world he does. There are no separation between authors and readers, both are users and contributors to the system. The Truth may stay the same, but how we see it will change. It has always changed, but it changes faster now. Being able to change with it is a competitive advantage.
Errors in the Encyclopædia Britannica which have been corrected in Wikipedia
While I don't disagree with all the points in this article, and thing the "trending towards mediocrity" issue is one that needs to be addressed (if you read the mailing list archives, it in fact has come up numerous times), Britannica is hardly a repository of flawless truth either.
For some examples from the other side, see:
Errors in Britannica which have been corrected in Wikipedia
10 PRINT CHR$(205.5+RND(1)); : GOTO 10