In the beginning was Wikipedia

It is a truth universally acknowledged that despite personal hard disks that can store more information than the average public library, new search technologies, multiple ways of organizing a lifetime of intellectual endeavor into files, and folders, and libraries, if you nevertheless set out to find one single known file on your laptop you will never, ever locate it.  You will be condemned to an eternity of playing the “You know, I’m sure I remember I had a file about. . . ” game.  Serendipity, however, still holds a powerful sway over our actions, and sometimes the fruitful search for what we knew once existed turns up something we had forgotten we’d ever had.

Such was the case recently when I was searching for an article I was sure that I had written a number of years ago.  As I was about to conclude that I must have dreamed it, I stumbled across an antique e-mail in my draft folder, a message that I’d written back in October of 2005 but never sent.  Why I never sent it I don’t remember.  Clearly it was an e-mail that was involved in a rather intense discussion so maybe I was doing that all-too-rare of human actions: hesitating before I actually sent an e-mail.  But since this is me, that is unlikely.  Maybe I felt the discussion had been beaten to death or wouldn’t be of interest to others.  Maybe I hadn’t finished the message, although it looks pretty complete to me now.

Whatever the reason, I was startled to find how closely the e-mail was reflecting many of my current reservations about the rise of the clowd.  This post itself doesn’t explicitly concern games.  But the Bioware/Hepler controversy that I wrote about in my last post has made me realize I need to pay some serious attention to one of the most pervasive myths of the technoosphere: that there is a wisdom in crowds.

Point
I think the conversation began when I posted a link to an article by Andrew Orlowski, an entrenched skeptic of Wikipedia.  “Wikipedia Founder Admits to Serious Quality Problems” was accompanied by the tag-line: “Yes, it’s garbage but it’s delivered so much faster.”  You would think that tells you all you need to know about Orlowski’s stance, and I seem to recall opening the e-mail exchange by describing this as a “gleeful cod-kicking” which it certainly is.  However, Orlowski makes some very good points, particularly concerning the degree to which Wikipedia received a remarkably free ride in both the mainstream and IT press in the early years.  And in the middle years.  And even in the later years.  Sure, problems have been pointed out, doubts voiced, suggestions for different processes raised.  But very few have questioned the basic premise of the project.

Counterpoint
I posted the link to the Orlowski article and probably accompanied it with numerous disparaging remarks about Wikipedia.  One of my colleagues, who was then engaged in an ongoing collaborative Wiki-based project with his first-year writing class, responded.

Mark et alia,

[WARNING: Long, and probably not worth it]

Let me preface this by quoting from Wikipedia co-founder Larry Sanger’s critique of what he sees as the anti-academic (shading toward anti-intellectual) stance of Wikipedia (see <http://www.kuro5hin.org/story/2004/12/30/142458/25&gt;) who writes:

“Let me preface this by saying that I know Wikipedia is very cool. A lot of people do not think so, but of course they are wrong.”

And before we start cod-kicking with too much glee, I think it’s worth stepping back (at the very least, that will give us better kicking leverage) to think about how open source/wiki projects work, not just in terms of generating more-or-less reliable and/or well-written informational product, but (more promising) in terms of the pedagogical process they model. In other words, to stay within the terms of generational response presented by Pete Townshend, I’m advocating an attitude more “Rough Boys” than “The Punk and the Godfather.”

I’ve been thinking about this because the students in my sections are currently revising the _University Writing and Intellectual Property_ web site that other sections have been working on for the past three semesters  a site that works as a (highly) modified wiki, in the sense that each semester, new students get to decide how they want to add to, revise, or renovate the site. And these students are quick to point out–and I’m ready to hear it–similar frustrations over quality to those outlined in the Andrew Orlowski article to which Mark refers us.

We’re at the point in our site’s life history where we might well ask something like Nicholas Carr does of Wikipedia in a blog entry from October 3, 2005 <http://www.roughtype.com/archives/2005/10/the_amorality_o.php&gt;:

“Remember, this emanation of collective intelligence is not just a couple of months old. It’s been around for nearly five years and has been worked over by many thousands of diligent contributors. At this point, it seems fair to ask exactly when the intelligence in ‘collective intelligence’ will begin to manifest itself. When will the great Wikipedia get good?”

And of course it is good, within limits: as Orlowski notes, good for “Klingon, BSD Unix, and Ayn Rand”; perhaps not so good (according to Larry Sanger) for “philosophy” or (according to Nicholas Carr) for “Bill Gates” or “Jane Fonda.” I’ll admit to using it myself quite a bit: it’s how I know the difference between “black metal” (mostly Norwegian) and “death metal” (mostly Swedish). But I’ll also admit that I rarely trust it’s usefulness on a topic to be determining in the last instance.

And yet, I hope we can agree that there is something pretty great about even the not so good, if we can get away from fetishizing the quality of the finished product, and look at the “culture of writing” that’s being modeled here: one based on broad peer review, accountability to a writing community, a habit/imperative of revision, and a habit/imperative of citation. (Sanger argues that the good faith implicit in all of these ideals quickly breaks down in practice, given the high troll-per-capita percentage of writers.)

My colleague then went on to discuss his own class project in more detail, ending on an interesting cautionary note where he voiced the concern that the real danger with Wikipedia wasn’t its lack of organization, but rather the fact that its organization would soon make it the target of attempts at organized manipulation by government and corporate entities.

Beside the Point?
I thought long and hard about some of the points my colleague had made and then worked up the response which I never sent.  This is what I wrote:

[Warning: probably even longer than ______’s post and probably not worth reading half as much as his. This post, like all my posts, also comeswith an NC 17 rating]

_____I love the way you manage to take my over-heated language and both point out how over-heated it is and give it a lot more intellectual respectability than it actually deserves! It sounds like you and your students are facing similar issues with your web project as the students are facing in my class, so I’ve got a couple of suggestions there. But we’re not done with Wikipedia yet. . .

Reading your thoughts on Wikipedia helped me clarify what I see as some of the major issues with the Wikipedia project as it stands at the moment. And this isn’t any reflection on the idea of Wiki projects in general, just this specific one. Wikipedia has achieved a near-mythic status in some quarters of the IT world (and as the Orlowski article pointed out, it’s had pretty much a free pass in much of the IT press until recently); it has also been the subject of fanatical devotion on the part of many Wikipediaphiles (sorry, couldn’t resist that one). So for me the question to ask is this: why is it that Wikipedia has received so little criticism from users, its authors, and the IT community? The short answer is that it represents some deeply held myths about the nature and direction of information technology, beliefs that are held in naive form by many non-IT specialists and in more sophisticated form by many working in the IT world.

The long answer breaks down something like this:

1) Wikipedia is great because it’s “open source.” Open source has become the mantra of those who advocate alternatives to the domination of the IT world by a few small players. And I also believe in the potential of open source. But at this point I think it’s worth asking whether or not Wikipedia is a true open source project (along the lines of Linux, for example). Open source projects have the aura of free-floating collaboration, but they are in fact based around very tightly controlled (and administered) protocols for error-checking, flow and version control, quality and need assessment, and the like. The protocols that Wikipedia has, such as they are, are quite patently inadequate in all these areas.

2) Wikipedia is great because it’s taking a world of knowledge and putting it online. Here is it probably worth asking a) whether or not the world needs another encyclopedia, b) if Wikipedia is doing anything in terms of its informational content that another encyclopedia isn’t doing, and c) if developing a massive encyclopedia is really exploiting the potential of collaboration and dispersion that the Web makes possible. And the answer to all three questions for me is a resounding, No.

3) Wikipedia is great because, unlike a traditional encyclopedia, all its information is cross-referenced. This, actually, is Wikipedia’s greatest weakness. Many entries are a mass of links. Check out this one on Plato that I pulled out at random: http://en.wikipedia.org/wiki/Plato Some of these links are useful for trying to understand Plato, but the vast proportion of them are irrelevant. Look, for example at the themes section where the following links take you to completely separate entries in the Wikipedia: democracy, monarchy, government, aristocracy, nature, convention, heredity, envirronment, intelligence, personality. . .I can’t be bothered typing the rest (and this is in three short paragraphs). There are two assumptions being made here with this elaborate link structure: 1) readers are stupid, and won’t be able to understand what Plato was all about unless they are able to define “government” or “personality” (an odd assumption for a project that places great stock in the notion of collective intelligence: many of its entries are written as if the expected user is a moron); 2) People looking for information on Plato in fact will need or desire to find information about everything else. Knowledge is specific, particular, and situated, but Wikipedia loves to pretend that it isn’t.

This is the fantasy, so beloved by technogiddies, of the “infoverse” with everything connected to everything else. The problem is, that is a fantasy only if you reduce knowledge to data and to the act of linking rather than understanding. At some level a lot of people understand that everything in the universe is connected to everything else. But that, functionally, doesn’t produce much in the way of useful knowledge that helps us get things done. So if Wikipedia is going to be regarded as something more than a vast, but ultimately tedious, piece of performance art demonstrating the Great Oneness of the Universe, it needs to use some of that collective intelligence to think long and hard about *motivated* linking rather than info-splicing.

4) Wikipedia is great because it has a lot of information. Unfortunately, as many of my students discovered (although they were certainly divided on this issue) it’s also impervious to some of the most basic fundamentals of web page design (the massive long pages, for example, or the fact that entries adopt an idiosyncractic arrangement of text and navigation boxes). So what Wikipedia fosters is a belief that “information” can be separated from the way in which that information is displayed. ______’s cautionary argument about this being a “tertiary” endeavour is a propos here, as is the piece my classes read in preparation for the web bibliography project, Geoffrey Nunberg’s “Farewell to the Information Age” where he analyzes the rise of the abstract sense of information, as against what he calls “intelligence” (information that is warranted in terms of source and is directed to some specific purpose).

5) Wikipedia is great because it’s a collaborative project. _____ makes this argument a lot more eloquently when he says: “And yet, I hope we can agree that there is something pretty great about even the not so good, if we can get away from fetishizing the quality of the finished product, and look at the “culture of writing” that’s being modeled here: one based on broad peer review, accountability to a writing community, a habit/imperative of revision, and a habit/imperative of citation.”  Myself, I don’t think it is “fetishizing” to demand that the project live up to its own claim to be providing (or working towards providing) a reasonably authoritative source (and that being authoritative also be understood to involve being credible in writing and presentation). If this were, after all, a big pedagogical project (and I agree with _____ that that is a great part of its utility) then there is absolutely no reason why the product needs to be available to anyone other than a subscriber base. More problematic, however, is that Wikipedia is modeled on a very naive belief that collaboration will automatically produce something that is superior to other ways of authoring, designing, and implementing projects. But it’s not necessarily true that the thousand monkeys will ever produce Hamlet.

Pointillism
So, an unbelievable seven years on from this exchange, what are we to make of Wikipedia now?  I would first of all say that my friend was right: there is something great about Wikipedia and I see a number of virtues in the project now that I didn’t at the time.  One of the ironies of Wikipedia is that it is often held up as the paradigmatic example of this new fast-paced age where we want quick information and we want it yesterday.  Yet, the actual process that leads to the creation of many of Wikipedia’s entries is the complete antithesis of our current zeitgeist.  It is cautious, deliberate, often painfully slow.  An article we often read in my classes is Tom Dunkel’s “Word War III” which describes the long involved process surrounding the growth of the Wikipedia entry on Mahmoud Ahmadinejad, a process so contentious that it led to the article being locked numerous times and resulted in several rounds of mediation.  But this process was unusual only in degree not in kind.  As Dunkel notes, Wikipedia operates according to “concensus time” and the editing process takes as long as it takes to achieve a workable consensus.  Therefore, the source to which we all turn because we are more concerned with efficient access to information rather than quality information is in fact the poster child for inefficiency.

Which makes it not unlike the scholarly peer review process.  Indeed, one of the reasons I ask my students to read the Dunkel piece is because it leads to a re-examination of the idea they have all had rammed into their heads in high school: that they must not ever, on pain of death, use Wikipedia in their papers because it is not scholarly.  The truth, however, is that it both is and is not a scholarly source.  It is certainly the ultimate form of peer review.  At this point my students usually object that the people editing Wikipedia articles are not “experts.”  The truth, of course, is that they simply aren’t recognized, credentialled experts.  Given the tendency of Wikipedia editing to be motivated by people who have a strong interest (often, it is true, an unhealthily strong interest) in specific topics, many of the entries are being edited and reviewed by experts.  Furthermore, when it comes to the scholarly process, because most forms of peer review are anonymous, we are in essence being asked to take it on faith that the people reviewing our work are experts.  I’ve had a lot of great peer review feedback over the years.  But I’m sure I’m not the only scholar who has also received feedback that seemed to be coming from the same Nigerian businessman who keeps e-mailing me about administering some funds in his possession. Wikipedia, furthermore, has one enormous advantage from the point of view of the average Joe or Jane: the entire review process, every painful edit and vitriolic flame war, is visible to the public.  The “real” scholarly process, by contrast, might as well be taking place in a smoky back room for all the public (or even the tiny subset of the public who represent the actual audience for our work) know.

So the process of Wikipedia has turned out to be, I would argue, one of the best examples of scholarly peer review our culture has ever known.  The problem, however, is that this great process of scholarly review is single-mindedly dedicated to producing material that isn’t scholarly.  Wikipedia shares the same flaw in this regard as all encyclopedias.  It is dedicated to status quo knowledge, to communicating the consensus view about various subjects.  Scholarly work, on the other hand, is (or should be) dedicated to producing new knowledge, a process that often involves challenging received wisdom and the consensus view of things.  Ironically, the vicious flame wars that often accompany the creation and editing of the more controversial Wikipedia entries are all in the service of eliminating any controversy from the resultant article.

Yet time has also proven both Orlowski and myself right.  One thing that Wikipedia is still wrestling with is the problem of overall quality control.  As Orlowski admits, reliability of some articles is excellent.  Yet most encyclopedia’s don’t measure their reliability based on the strongest entries but on that of their weakest (or less prominent) ones.  Perhaps the greatest challenge faced by Wikipedia is not reliability but readability.  Orlowski notes that ” Even when a Wikipedia entry is 100 per cent factually correct, and those facts have been carefully chosen, it all too often reads as if it has been translated from one language to another then into to a third, passing an illiterate translator at each stage.”  This is one of the major reasons why consulting Wikipedia still very often leaves me feeling as if I need to take a bath.  I am fairly confident this will, however, become less of a concern over time.  Not because Wikipedia will get any better in this regard, but because Twitter and Facebook will amplify the indecencies already perpetrated upon language by politicians, sports casters and journalists, until there won’t be many people who give a rat’s ass about style.  Anyone timidly standing up for eloquence will be put in the same category as someone bemoaning the loss of papyrus as a medium.

In hindsight Orlowski was also right to raise questions about Wikipedia’s process.  Wikipedia itself has backpedaled. . .er, staged a strategic withdrawal. . .er modified the core processes of the site in numerous ways in recent years.  Restrictions have been placed on who can edit certain kinds of articles (chiefly those for whom people are still alive), a more comprehensive moderation and arbitration process is in place (as described in Dunkel’s piece) and the site now aggressively warns users about the reliability of its information with a layered series of labels.  The organization has now begun to actively recruit editors rather than relying entirely on happy volunteer accidents.

I also think I was right about many of the elements I noticed way back when.  Wikipedia was the way it was and still is the way it is mainly because it represents several core beliefs concerning the role of IT in our world: beliefs about the utility of collaboration, of collective wisdom, of a massively overlinked multiverse of information, and the sacrosanct nature of open processes.  Far from being questioned, these assumptions now underlie numerous cultural processes from Facebook to videogames.

In addition to the elements that I discussed then, one additional IT canard that informed the early stages of the Wikipedia project has been de facto called into question.  Everyone has heard the old hacker slogan “Information wants to be free.”  But it turns out that information doesn’t want to be free: instead, it wants you to buy it an Alexander McQueen dress and fly it to Paris for a night on the town.  Information costs.  Good information costs even more.  So the one thing that is new about the new Wikipedia is that every time you go to the site now there is a good chance you will be hit up for money.  Wikipedia has discovered what US Public Radio has known for a long time: that if you want to step outside the commercial paradigm, if you don’t want to treat information as a product but rather as a service, you are going to be reliant on the kindness of strangers.  The organization is also simply (re)discovering what many in IT support have known for years: that promoting innovation and even getting it off to a flying start is actually the easy part: it is keeping that innovation going that is the hard part.  The hardware that stores and backs up and backs up again and transmits all this “free” information costs.  Quality control costs.  Wikipedia, like most other significant collaborative innovations of our time, is having to figure out how to pay for its ongoing innovation.  And there aren’t too many options.  You can be like Zukerberg and whore your collaborators out to the highest bidder.  Or you can hit the creators up for donations.  Or you start charging for your product just like a regular business.

Unfortunately, none of these lessons seem to have been learned.  So we are being marched into a brave new future by those convinced of the magical self-organizing powers inherent in the clash of wannabe free information and the collective intelligence of mobs, all supported by a mystically self-supporting hardware infrastructure.  And lo, the clowd is born.