I hope that this makes for suitable "Meta" discussion. The source comes from the Foundation-l mailing list, a post by Ray Saintonge:



This neatly sums up the anti-expert culture that pervades Wikipedia, along with the blind notion that if "trust" is extended to everyone (even paid teams of content manipulators), everything will eventually work out okay.

There's a grain of truth here, I think, but it's being distorted. More than anything I have ever seen, Wikipedia is "fundamentally based on... tradition". Because of the way the so-called consensus model works, nothing ever changes on Wikipedia. Even the most obviously beneficial changes are forever stonewalled - it makes the Senate look like a body that moves too quickly.

It is hierarchic organizational patterns that allow for swift and needed changes. That's why totalitarian regimes can accomplish (albeit at great cost) things that no democracy ever could, and that's why the centralized modern state came to dominate as a form of political organization, sweeping away feudalism and confederation.

If Wikipedia worked the way that proponents would have you think, with more nimbleness and dynamism than its rivals, then it might well be an incredible thing. There is a certain micro-level dynamism on Wikipedia. It is truly remarkable how quickly articles can be written on current events, for example. The problem, though, is that this micro-dynamism is paired with an intense status quoism at the macro-level. Whatever dynamism the wiki-model creates at the micro-level is nearly destroyed by the macro-level difficulties.

What this reminds me of is the whole debate over 4th generation warfare/netwar. Decentralized, networked organizations have proven incredibly effective as a form of insurgency. You can inflict a lot of damage and even defeat your opponents through networked organization, but it doesn't work as a form of governance. Once the shooting stops, you must adopt a more conventional approach to "win the peace".

This is why George Washington was an important figure. Tactically and operationally speaking, Washington was very close to being a disaster. In his whole career, he only "won" two engagements (Trenton and Princeton), and if the fate of the American colonies had ridden on his battlefield skills, then independence would never have been won. It's the people like Nathaniel Greene who won the tactical and operational victories by fighting in a way quite unlike the British while Washington was, it seemed, wasting time and resources trying to fight traditional engagements.

So why does Washington matter? Because he recognized on a strategic level that the Americans needed to do more than just win battles. They needed to create a suitable political-military organization to compete on level terms with the British in order to establish a viable state. Strategists of insurgency ever since have recognized the same. Thus, for example, Tet (while a tactical and operational disaster) was the greatest strategic victory in the whole war for the VC and the PAVN.

Anyway, sorry for the detour in strategic thought, but the point I am trying to make is this. Wikipedia has found an effective tactical-level method of engagement, that is frankly superior to that of its opponents. On the basis of this strength at the tactical level, Wikipedia has succeeded on many metrics (traffic, breadth of coverage, etc.) but it is still a failure on the grand-strategic plane. In order to achieve success there, Wikipedia must improve its organizational model, and the lesson of past insurgencies, is that in so doing it must much more closely approximate the models it seeks to reject. A hierarchic governance structure is simply the best way forward for an endeavor of Wikipedia's size.

The problem is many Wikipedians are in love with the atomized content and market simulation aspects of wikis which are more of a problem than bad management or a stupid godking. Substitute "someone being responsible" for "hierarchic governance." The reason insurgencies need to be decentralized is so the regimen's forces have no clear target to return attacks. Great for asymmetric warfare. Not so good if you are trying to address defamation or protect children.

Yeh. This sounds like content that should be discussed on WP:Fringe theories/Noticeboard. The New Age dawns on Wikipedia.

While it may not be "respect and tradition" that is needed (try expecting respect and see how far it gets you), the consensus model just doesn't scale. So something else is needed. Wikipedia isn't an experiment in government, so it's not necessarily "governance" but some kind of management is.

I think there's a large number of folk in the en:wp community who either don't understand this, or do understand it, but choose to make use of that fact to keep the status quo in place for whatever reason.

Wikipedia does not have and never had a "consensus" model of decision making. I don't mean they have a funny version of "consensus" either. Inherent in instant atomized content creation is an ultra individualist principle that mimics market decision making. The wiki "decides" to say whatever the last editor chooses that it says. This is suppose to result in self improving content. All the various "consensus" layers exist because the primary market like model doesn't actually work.

There is nothing "market like" in the primary WP model of what they call "consensus." A "market" by definition is a venue with open and free trade ("free" doesn't mean that theft is allowed, but it does presume no draconian tariffs or import/export quotas and the like). It really does operate by a kind of consensus, which is what sets a SINGLE price. You cannot have a single price at any given time, without consensus as to what that price should be.

Whenever you have one trading venue insulated or semi-insulated from another, by tariffs or quotas or local controls, or the sorts of things that make the same pharmaceuticals 50% more expensive in the US than in Canada or the UK (or the rest of the Western world, really), then you don't have ONE market. You have two or more markets. TWO greatly differing prices for the same good DEFINES two markets.

Now, compare this to the thoroughly atomized (all right, Balkanized) "consensus" process on WP, in which things are decided for each article among small groups of local warlords and power-owners, who throroughly (and effectively) resist having any standards for decission-making cross local article-boundaries. By using little mantras like OTHERSTUFFEXISTs to keep standards in one argument about one article, from being applied to any other argument about any other article.

There's nothing market-like about this at all. It's a thousand or a million little "markets" (warlord territories), each with their own separate "consensus" process, and some minor smuggling that connects them.

It is "market-like" in that:

• It involves completely individual choices;
• It is controlled by the "invisible hand "of the aggregate of many supposedly self correcting interactions;
• It produces the illusion of a complete lack of coercion;
• It doesn't really work, except to reinforce the position of pre-established elites.

Consensus is not "atomized." It is an inherently collective process and deal with issues as whole pieces, even in the distorted form used on Wikipedia.

The things you say about warlords and all are more or less correct. The constitute the sick and distorted political economy of Wikipedia. Of course one of the purposes of "markets" is so that societies can deny they even have political economy and to hide it from view.

And what was it about WP that gave you the illusion of complete lack of coercion?

I think the rest of us missed that illusion.

Was it good for you?

When I am at the newsstand at the train station, I can purchase a twin 2-pill pack of acetaminophen for $2.49. That's about 63 cents per 500 mg tablet. Or, I can walk about 600 feet down the corridor to the Dollar Tree and buy a 50-pill bottle of acetaminophen for... wait for it... $1.00. That is exactly 2 cents per 500 mg tablet. The only visible difference is that the newsstand brand is one I've heard of (Tylenol), while the Dollar Tree is selling a brand called "Assured". I am one of those consumers who feel that something as common as acetaminophen is pretty much a commodity, and so I don't choose on the basis of brand name.

So, the 31-fold difference in pricing along this corridor at the SEPTA Suburban Station in Philadelphia is explained by insulation between the stores? Tariffs? Local controls?

I had no idea there were so many "markets" in Suburban Station.

Exactly right. Wikipedia likes to advertise itself as a "free market of ideas", when in fact it's just as feudal and Balkanized as Europe in the 10th century. The rub comes in with the way they keep it Balkanized -- in secret, and with secrets that are very badly kept as well. A masterpiece of mismanagement, where the mismanagement is the model, and nobody admits or even believes there is mismanagement.

It's too bad Marshall McLuhan didn't live to see Wikipedia. He would have had a field day trying to explain it.

You're buying two different goods. One is called "acetaminophen" and the other "acetaminophen+600 feet walk". Product differentiation (brand names), transaction costs (like having to walk) etc. can give rise to a bit of monopoly power - the situation is usually referred to as "monopolistic competition". In fact, things like tariffs, local control, etc. that Milton brings up are essentially "artificial transaction costs", created by law or decree. But they tend to have the same effect as naturally occuring transaction costs. (There's a famous quote by quasi-Marxist economist Joan Robinson about tariffs having exactly the same effect as throwing big rocks into your harbor)

Discussing the ins and outs would probably lead into areas irrelevant for the present discussion about Wikipedia



As I've mentioned in the other thread, to the extent that this is a valid analogy, what makes markets work well is competition. So articles with lots of watchers, or to be more precise lots of watchers with different opinions and views are going to end up half decent, while those that only a particular clique cares about are going to be shite - like a monopoly situation. A frequent problem that arises is in between - there's two groups that care about a particular issue (while rest of the Wikipedia thinks they're both crazies) and they're basically locked in a struggle - an oligopoly.

The thing is, competitive markets cannot function or even arise without good institutions to back them up (i.e. "missing markets", "incomplete markets" etc. - only extreme libertarians or anarchist would argue that markets can function without these, or that good institutions will "naturally" spring up). And just like crappy "institutions" often create monopolies in the real world ("unnatural monopolies"), stupid Wikipedia policies encourage the creation of medieval fiefdoms and the like.

That generic meds rarely retail in serving-size (i.e. a non-bulk quantity appropriate to the acute ailment of one person) sure doesn't help anything either.

Rather it encourages (a) buying name-brands at highway-robbery price, (b) over-medicating due to a surplus behind the mirror, or (c) wasting most of the stuff.

And this illustrates my problem with using market language to talk about Wikipedia: the theoretical market that armchair economists like to talk about does not exist, has never existed, and cannot exist. No real market comes sufficiently close to an ideal market for ideal market rules to meaningfully hold; this is no more or less true in Wikipedia than it is every else.

Let's just admit that trying to model social behavior within Wikipedia using market models is doomed to failure, and move on.

I have big reservations about using the market analogy for Wikipedia as well. There's probably some way to model Wikipedia as some kind of market or other similar institutions but "competitive market" isn't close by any means.

By the way, the purpose of a theoretical construct such as a competitive market by economists is not to reflect reality 100% accurately - of course generally these don't really exist - in the same that a purpose of a road map is not to model every aspect of physical reality 100% accurately - and such a map would be infeasible and/or useless. (yes, this is an overused metaphor)

Yep. But I'm going to riff a bit on it anyway. At a train station you can get away with a pretty high transaction cost for a 600 ft walk, if the last train may be arriving in a time frame that prevents you from making the longer walk, and what you're paying for is half an hour less of headache, while you ride that train. At convenience locations you pay for convenience. Basically you pay for not having had presense of mind to bring your own.

The brand label is an interesting thing, too. You pay for two different services there. One is the assurance of quality by the brand-labeler, as well as the assurance that you're getting the product you've seen advertised (maybe you're not perfectly sure that acetaminophen is the same thing as Tylenol-- even if you have your smart-phone, how long is it going to take to convince yourself of this, and how much is that call going to cost you?).

The other thing is an intangible sense of satisfaction that many people get from a brand-name label. They'd get that even if they KNEW that the product came from the same factory, and somebody had just labeled it differently. Or (as is more likely the case) even if they knew that blinded experts could not tell the difference, or even chose the less expensive product as being better.

Strange things happened in the wine industry some years ago when wine taste tests were finally forced to be done blind. California wines began beat French wines, sometimes. Sometimes FAR more expensive French wines.

So -- what exactly is it that you buy, when you buy a French wine? Some of it is the label. You buy WINE. FROM. FRANCE. You pay for the assurance. Knowing the provenance may even make it taste better to you. It's a sort of placebo effect. You're paying for the release of your own endorphins; pleasure-chemicals made by your own brain. You'd like to release them in your brain by some less expensive means, but the damned things won't cooperate! Such is the nature of buzz.

Milton

Yeah, it was a crazy post, and when I read it I have to wonder if maybe it was actually intended to be ironic. I mean, it's basically a proof by reductio ad absurdium against what it calls "the essence of wikiness".