Wikipedia and the Authority of Collective Intelligence

Introduction

It has been described as the Font of All Knowledge (Berners-Lee 2004). It contains over 1.5 million articles and is written in 120 different languages (Zachte 2005). The talk is of Wikipedia, an online Encyclopedia that gives each user the ability to create and edit its contents. This simple idea of being ‘from the people’ has brought up much controversy about the accuracy of its information. This report compares three different standpoints upon the topic of Wikipedia’s authority and the authority of collective intelligence. Correspondingly, this report makes a point on when to, or not to, use Wikipedia as a source of information.

The 3 authors of the articles in comparison are Peter Morville, John Seigenthaler and Nicholas Carr. Each author comes from a different corner of the industry. One comes from politics, one from information architecture and one from business. While each author has a different opinion upon the topic, they all have some things in common. They are all good writers, both in journalism and in publishing. They are all masters of their field. And they all use Wikipedia.

Up front, Morville puts the aspect of collective intelligence in a better light than his colleges. We will see, during the course of this report, if his arguments hold up against Seigenthaler’s and Carr’s.

Article 1: Authority

About the Author

Peter Morville has a father role in the information architecture field. He is the president of Semantic Studios, a company that specializes in information architecture, user experience, and findability. Morville is the co-author of the best-selling book “Information Architecture for the World Wide Web”. He frequently publishes articles on his companies web site.

Critical Analysis

Morville’s article captures the readers interest from the beginning where he admits: “I have a problem with authority“ (2005). From there, he dives into the topic of authority, looking back on how authority was, then describing how it has become. He states, in my opinion correctly, that the way we handle authority has changed immensely in the recent past. In modern research, we do not only have to find information, but we must also have the ability to judge upon its correctness. The Wikipedia is a big help because its information is accessible, yet gives us the new problem of judging upon it. Morville states that, based on three topics he looked up, the accuracy is quite good. I agree some topics are of good quality, yet miss that he does not give a ‘bad’ example to reinforce his statement. As we will see in the next articles, the bad examples are numerous. Morville does a fair job in comparing the Wikipedia to the old school of libraries and librarians, tagging some of these people as nay-sayers.

Morville’s defines five factors for measuring authority:

”Authority derives from the information architecture, visual design, governance, and brand of the Wikipedia, and from widespread faith in intellectual honesty and the power of collective intelligence.”

I will now set my focus upon these five factors.

Information Architecture

The findability and ease of use of Wikipedia boosts its credibility. Meaning for example, if a Wikipedia entry is found and read more and referenced to more than another article, it in time is perceived as more reliable. As Morville states:

”Wikipedia doesn’t beat Britannica because it has better folksonomies. It wins because it’s more findable.”

Looking at the issue the other way around: If information cannot be found and nobody has access to it, then the information is of no use, even if it is fully reliable. Wikipedia wins in this aspect, and yes I agree, Information Architecture plays a role in the definition of authority. Good information architecture results in higher popularity of the information. I will discuss popularity later in my conclusions.

Visual Design

According to Donald Norman, author of the book “Emotional Design“, information is perceived as ‘good‘ when it looks that way (2004). One example in Norman’s book shows:

”Attractive things make people feel good, which in turn makes them think more creatively. How does that make something easier to use? Simple, by making it easier for people to find solutions to the problems they encounter.”

By triggering emotions with design, problems are solved easier. The presentation of information cannot compensate for sloppy information but makes accurate information better. Wikipedia presents its information in a clear manner. Morville is correct in his statement, and so Wikipedia wins again, even though I personally would not weigh visual design heavily when measuring authority.

Brand

Encyclopedias and dictionaries are branded. When I think of the name “Webster“ I automatically think of a renown encyclopedia. I also trust that Webster is a good source of information. Wikipedia is a brand name, but competing with other names, such as “Webster” or “Encarta”, it has less authority in my own opinion. Again, I confirm Morville’s statement that information authority is partly derived from a brand.

Intellectual Honesty

While all the other factors Morville defines for authority are logical and therefor almost untouched by critique. The ‘collective intelligence’ front of Wikipedia is attacked the hardest. By putting faith into collective intelligence, Morville makes his strongest point in his articles argument. In his opinion, the more people are in charge of content, the better the content will become. Collective intelligence is not single sourced and therefore not biased. I share his view, especially in news and politics, multiple perspectives of the same story are relevant. This aspect is a major plus over classic media. The downside to collective intelligence and the Wikipedia, as Morville writes, are the amateurs and trolls that post false information. Will this group be the downfall of the Wikipedia? Not if there is enough governance.

Governance

Wikipedia leaves the governance up to volunteers, as Seigenthaler has found out by asking Jimbo Wales, the founder of Wikipedia:

”Wikipedia’s web site acknowledges that it is not responsible for inaccurate information, but Wales […] insisted that his web site is accountable and that his community of thousands of volunteer editors (he said he has only one paid employee) corrects mistakes within minutes.”

Governance is a serious issue for Wikipedia. There is some basic form of governance of the Wikipedia entries, but the overall amount of governance is lacking. We will see what the lack of governance can lead to in the next article.

Article 2: A false Wikipedia ‘biography’

About the Author

John Seigenthaler is a journalist, a writer, and a political figure. He is the founder of the Freedom Forum First Amendment Center at Vanderbilt University and also served as president of the American Society of Newspaper Editors. Seigenthaler is now in retirement and occasionally writes for the editorial page of USA TODAY.

Critical Analysis

As the founder of the First Amendment Center, Seigenthaler is an advocate of the freedom of speech. Ironically Seigenthaler’s image is being hurt by an advanced form of this freedom.

Seigenthaler’s standpoint on Wikipedia’s authority is in high contrast to Morville’s. For a good reason. While Morville is personally untouched by false information, Seigenthaler seems to be hurt. When reading his article, I got a good sense of Seigenthaler’s frustration. His writing is direct, simple and to the point. Seigenthaler is dealing with the direct misuse of information. The author of his Wikipedia entry, called a ‘troll’ in web jargon, inserted lies about his life. While fighting back, Seigenthaler makes some interesting findings. Even with his social status and his power, he is not able to find the author of his biography. Proving that editors on Wikipedia are almost anonymous to its readers. Another finding is that the Wikipedia information is reused on other web sites, that lag behind with updating. Proving that Wikipedia has a far reach. Luckily this reach is only temporary. A Wikipedia entry can be revised instantaneously, unlike a printed medium, reducing the amount damage. Seigenthaler, with his own authority, was able to delete the history of his Wikipedia entry.

Anonymity

In my mind, Seigenthaler’s strongest point in his argumentation is that Wikipedia editors are not only anonymous (with the exception of an Internet Protocol Address) but also that their privacy is protected by law (Fusco 2000). As of yet, posting false information returns no direct effect to the author. It is frightening to know that one could re-edit Seigenthaler’s entry and probably, if careful enough, get away with it. Yet, for collective intelligence, being anonymous is not only a bad thing. Blinded by his ‘rage’, Seigenthaler does not comment on any good aspects. He does not give in to any plus points, which makes his article one-sided and not as balanced as a Morville’s and Carr’s articles. Due mainly to the fact that he is writing his story himself, and not is being written about, by a neutral journalist. The good side of the anonymity of collective intelligence, is that it allows people to express their views without having to be afraid of a competitor. A strong plus in my earlier example of news and politics (in the analysis of Morville’s article).

I like Seigenthaler’s articles ending, using a fundamental teaching he got as a child and projecting it onto Wikipedia’s concept.

When I was a child, my mother lectured me on the evils of “gossip.” She held a feather pillow and said, “If I tear this open, the feathers will fly to the four winds, and I could never get them back in the pillow.”

The metaphor imposes that the flying feathers (the misinformation) can never be undone and returned to the pillow (the source). Updating the metaphor for the online medium, one would have to include the possibility to switch the feathers in mid flight (by correcting the false information).

Time

The fast nature of Wikipedia is already described in its name. “Wiki” is Hawaiian for “quick“. As Seigenthaler writes, he did correct his entry as soon as possible, yet the following remains a fact:

“For four months, Wikipedia depicted me as a suspected assassin before Wales erased it from his web site’s history Oct. 5. The falsehoods remained on Answers.com and Reference.com for three more weeks.“

Considering the factor of time and that information gets changed with a fast pace, I think that a Wikipedia article should never be directly referenced to. For it is not a certainty that someone following that reference will be served with the same knowledge. An exception could be in a presentation, where the entry was checked shortly before.

Accountability

10 days after publishing the article on USA-Today, the publisher of the false Wikipedia information showed up in Seigenthaler’s office confessing what he had done (Seelye 2005). Upon this Seigenthaler states: ”I still believe in free expression, what I want is accountability” (2005). Seigenthaler wishes that Wikipedia editors can be held accountable for their entries. While I agree to the extent that accountability should be a measurement of information authority, I will argue that authority also means the loss of anonymity later in the conclusions. It is interesting to note that the ‘troll‘, named Chase, lost his job as a consequence to his posting.

The reason why Chase even made the entry is that he thought Wikipedia was a joke, as Katharine Seelye of the New York Times writes:

“In a confessional letter to Mr. Seigenthaler, Mr. Chase said he thought Wikipedia was a ‘gag’ Web site and that he had written the assassination tale to shock a co-worker, who knew of the Seigenthaler family and its illustrious history in Nashville.“

The ‘stupidity‘ of some Wikipedia authors is covered by Carr in the next article.

Article 3: The Law Of The Wiki

About the Author

Nicholas Carr is an acclaimed business writer and speaker whose work centers on strategy, innovation, and technology. After being the executive editor of the Harvard Business Review, Carr wrote the book “Does IT Matter?“, that focuses on the role of computers in business. Aside from writing articles for well known business newspapers and web sites, Carr posts his views on his web log named “Rough Type”.

Critical Analysis

Carr opens his article by first citing another journalist. This is an interesting method: He first looks back to what has been said before and then builds upon it. His writing style is casual, incorporating even the use of vulgar language to enhance his point. I have the impression that, for his personal articles, Carr breaks free from his otherwise strict business approach. This does not mean that his arguments are less powerful. Carr went out on a limb in a previous article titled “The amorality of Web 2.0“ (2005b), by briefly yet strongly criticizing Wikipedia’s authority:

“In theory, Wikipedia is a beautiful thing - it has to be a beautiful thing if the Web is leading us to a higher consciousness. In reality, though, Wikipedia isn’t very good at all. Certainly, it’s useful - I regularly consult it to get a quick gloss on a subject. But at a factual level it’s unreliable, and the writing is often appalling. I wouldn’t depend on it as a source, and I certainly wouldn’t recommend it to a student writing a research paper.“

This article, “The Law Of The Wiki“ is a brief follow-up to his initial statement, with more fundament.

General-Interest Content

Carr explains that general-interest topics in Wikipedia have no authority by using the examples of biographies. I’m sure Seigenthaler can agree to this, after what happened with his biography. Carr creates a ‘Law of the Wiki‘, in which he states: “Output quality declines as the number of contributors increases“ (2005). With this statement Carr tramps on the weakness of Morville’s main argument that collective intelligence makes content better, by guessing that the amount of amateurs and trolls are a large percentage of the group. Carr foresees that the weight is shifting toward stupidity and away from intelligence. I can not yet agree with Carr’s theory, unless he could prove that the overall quality of the Wikipedia is worsening. Yet, he does narrow down his rule to the general-interest subjects. And states that, on the contrary, the specialized content is quite good.

Specialized Content

“Nature”, an international weekly journal, tested the authority of Wikipedia’s specialized content in their December 15 2005 Issue. Nature compared 42 topics between Wikipedia and Britannica by giving them to acknowledged experts and hiding its source. I find the outcome of this test somewhat surprising:

“The exercise revealed numerous errors in both encyclopedias, but among 42 entries tested, the difference in accuracy was not great: the average science entry in Wikipedia contained around four inaccuracies; Britannica, around three.”

I am surprised about two things in the above statement. First, Britannica to me has such authority, that I would have awaited zero inaccuracies. Second, Wikipedia is close behind with just one inaccuracy less. Nature proves arguments from both Carr and Morville, that Wikipedia’s specialized content is generally reliable.

Centralized Control

As a solution to make Wikipedia more credible, Carr pleas for more governance. He proposes a centralized control mechanism, like the ones found in open source software projects. Meaning, someone judges upon an entries authority before publishing it, instead of making it ‘live‘ instantaneously. This idea is also found in the Wiki’s closest relative: the forum. In popular forum software, such as “phpBB”, it is possible to create administrators in the topic hierarchy, that are aloud to delete or change entries. These administrators have a function similar to the police in the real world. Carr believes in such an approach by writing:

“If Wikipedia wants to achieve it’s goal of being authoritative, I think it will have to abandon its current structure, admit that ‘collective intelligence’ makes a pretty buzz phrase but a poor organizational model, and define and impose some kind of hierarchical power structure.”

This leads us to a fork in the road. To the right, a sign points us to a policed, controlled Wikipedia with more authority. The sign on the left leads us to a pure democracy of information, with the possibility that the information is incorrect. Acting upon his idea of governance, Carr would take the path on the right. Morville, after making his view clear with phrases like: “The Revolution is Multi-Algorithmic” (2005) and: “The truth is a virus of many colors” (2005), would take the path on the left. I would follow Morville, arguing that there are yet more possibilities to assess authority, as I will now focus upon in my conclusions.

Conclusions

While analyzing the three authors articles, I commented upon each their findings of what makes information authoritative. Yet there are more mechanisms:

Referencing

Tim Berners-Lee, the initial creator of the web as we know it, was asked in an interview about the current state of information correctness. He replied:

”[…] When you go onto the internet, if you really rummage around randomly then how do you hope to find something of any of value? But when you use the web, you follow links and you should keep bookmarks of the places where following links turns out to be a good idea. When you go to a site and it gives you pointers to places that you find are horrible or unreliable, then don’t go there again. You see out there right now, for example, when you look at bloggers some of them are very careful. A good blogger when he says that something’s happened will have a point to back, and there’s a certain ethos within the blogging community, you always point to your source, you point all the way back to the original article. If you’re looking at something and you don’t know where it comes from, if there’s no pointer to the source, you can ignore it.”

In his answer, Berners-Lee points out that sites that are not perceived as credible are not re-visited, including sites with links to unreliable sites. Then, using an example of blogging, Berners-Lee stresses the importance of referencing. Thus, checking a Wikipedia’s credibility can be based upon its referencing. And if there is no referencing at all, the notion of the credibility of the entry drops.

Popularity

While Morville does not define popularity directly in his criteria of authority he does make a link to it with findability. In the ‘Google Economy‘ as Morville puts it, information becomes popular when it is found more and therefore is perceived to contain more truth.

Jason Thomale, a librarian of Texas Tech university shares this view by commenting:

“As librarians, we very much fight against this notion that a piece of information’s veracity depends on its popularity (or accessibility, or findability, or whatever word you want to use). That’s why so many librarians have reservations about using Wikipedia as a sole, authoritative source of information – because it has no authority other than what ‘we the people’ have given it. As librarians, we believe in some semblance of objective authority. We believe in evaluating information more or less on its own merits […].“

Thomale suggests, that when evaluating the quality of information, it’s popularity has to be taken into account. He also states that librarians (the predecessors of information architects), share a different view upon credibility and do not measure authority by including its usage.

Accountability

Seigenthaler wants accountability for Wikipedia entries. A good point. But accountability, by introducing a sort of registration system to Wikipedia, could result in the loss of anonymity. I would therefore suggest to Wikipedia to impose user-registration but only publish the authors avatar openly. The true information such as name and contact details should be kept secret by Wikipedia, following an online privacy policy which is verified by a separate company such as “TRUSTe”. Wikipedia would then have more power over its authors while at the same time protecting their identities from the outside. User-registration is a measure I would be willing to follow in order to keep Wikipedia authoritative.

Participation

One of the outcomes of the discussion that took place after Carr’s article in the comment-section, is that of participation (Carr et al. 2005). When you find inaccurate or insufficient information, don’t just move on, work with it. The point of collective intelligence is that people contribute intelligence. The discussion differentiated between writers and experts. The writer Jud wrapped up the conclusion nicely by commenting:

“And that’s the point I think everybody is making. Let the writers and editors write and edit. Let the experts contribute expertise. Let those who can do both be praised throughout the land.“

In other words, participate with what you are good at. This will make the content better, and therefore more authoritative, over time.

Recapitulation

In this article, I have presented the main criteria of what makes modern information reliable. While Wikipedia meets most of these criteria, some aspects need to be worked on. As of 2005, Wikipedia is a powerful source of information, if used correctly. As I have discussed, the following rules apply:

Overall, by putting faith into the power of collective intelligence, I am optimistic about the Wikipedia and its future.

In a personal appeal the founder of Wikipedia, Jimmy Wales, has called out for financial some financial help to keep the ever expanding Wikipedia afloat (without having to use advertising). I have not yet donated anything in 2006, so the free encyclopedia is definitely a good place to start.