This is a follow-up on something I posted a few weeks ago, with comments from Stu, Ian and TJ. The issue was about moderators becoming some sort of gatekeepers over time and the existence of some sort of continuum across which this happens.
While researching some of the control issues related to Wikipedia, I happened upon some very useful comments from Clay Shirky on all this; he was in turn responding to something Nicholas Carr had written on the “death” of Wikipedia; by following the Shirky link, you should also be able to see Carr’s original piece and his rejoinder to Shirky.
A few quotes from the Shirky piece:
- Openness allows for innovation. Innovation creates value. Value creates incentive. If that were all there was, it would be a virtuous circle, because the incentive would be to create more value. But incentive is value-neutral, so it also creates distortions — free riders, attempts to protect value by stifling competition, and so on. And distortions threaten openness.
- As a result, successful open systems create the very conditions that require [a response that threatens] openness. Systems that handle this pressure effectively continue (Slashdot comments.) Systems that can’t or don’t find ways to balance openness and closedness — to become semi-protected — fail (Usenet.)
- A huge number of our current systems are hanging in the balance, because the more valuable a system, the greater the incentive for free-riding. Our largest and most spontaneous sources of conversation and collaboration are busily being retrofit with filters and logins and distributed ID systems, in an attempt to save some of what is good about openness while defending against Wiki spam, email spam, comment spam, splogs, and other attempts at free-riding. Wikipedia falls into that category.
- And this is the possibility that Carr doesn’t entertain, but is implicit in his earlier work — this isn’t happening because the Wikipedia model is a failure, it is happening because it is a success.
Shirky notes later on in the same piece that “the rise of governance models is a reaction to the success that creates incentives to vandalism and other forms of attack or distortion.“
In the context of the continuum put forward by Ian and TJ, I think this is an important statement, and one that bears further discussion and analysis.
We need to think of governance models as evolutionary responses rather than immune system responses. An evolutionary response adapts to changing external stimuli in order to preserve and extend existence. An immune system response seeks to annihilate the interloper. While both are defensive in nature, I believe the principles by which they operate are as far apart as is possible.
An evolutionary response is open and knows no taboos. An immune system response is closed and knows only taboos.
We need to ensure that as we move across this suggested continuum, we stay on an open-with-as-few-controls-as-are required approach, rather than a this-doesn’t-work-so-let’s-bury-it-in-six-feet-of-concrete approach. What I have seen so far is too often the latter, which is why I brought this up again.
I think this is a case where I would like to tap into your wisdom as an economist. My blog entry for August 4 was written at a time when Wikipedia was taking a beating from both THE ONION and Stephen Colbert. I was wondering whether or not Wikipedia might experience an information-based variation on Gresham’s Law, in the sense that content that is “artificial” (made up without necessarily being warranted) could drive out more substantiated content. You are probably in a better position to assess the validity of this potential analogy than I am.
If there IS an “information version” of Gresham’s Law, then there is probably also a corollary similar to the observation that an “artificial” currency, such as paper money, receives its substantiation through the authority of governance. The same could be said of “artificial content.” It only has value if one can still count on it being warranted, and governance is one way to assure that warranting.
Here are two relevant blog entries. The first (which motivated this comment) is from August 4:
http://blog.360.yahoo.com/blog-Mff23hgidqmHGqbcv.lfskakEtS6qLVHUEMFUG4-?cq=1&p=64
The other concerns news that the German Wikipedia will be running an experiement in governance-by-editors:
http://blog.360.yahoo.com/blog-Mff23hgidqmHGqbcv.lfskakEtS6qLVHUEMFUG4-?cq=1&p=81
Agree totally. The need for governance models (emergent, opensource etc) is pressing.
It’s not just vandalism and free-riding in open systems I am concerned having just read the debate about Enterprise 2.0 here:
http://en.wikipedia.org/wiki/Wikipedia:Articles_for_deletion/Enterprise_2.0_%28second_nomination%29
via this: http://blog.hbs.edu/faculty/amcafee/index.php/faculty_amcafee_v3/the_get_out_the_vote_drive1/
What disturbed me about the debate was the constant reference by the editors (perhaps the ones resposible for deleting the original Enterprise 2.0 wikipedia entry) to The Wikipedia Policy. Their responses had that familiar authoritarian overtone, so beloved of corporate gatekeepers everywhere. What happens if the emergent protocol for something that works gets distorted in the name of protecting Wikipedia’s intergrity?
I am sure that there are parallels here with case/precedent law. Reading the comments I found myself looking for examples of wikipedia entries that equally didn’t satisfy the criteria on which the editors were insisting for Enterprise 2.0 entry. Take the entry Moonbat (http://en.wikipedia.org/wiki/Moonbat), for example…
1) neologism
2) only references on blogs
3) no primary academic or verifiable reference, let alone secondary…
Go figure.
I figure that it was the personal biases evident in the debate (and I am not referring to the defenders of the entry!) that fuelled the debate.
The danger of open systems that work and attract distortion (gatekeepers, control, vandalism) is summed up by a commenter on one of the blog posts about the debate: Ah, the beauty of empowering people with no other outlet of power.
http://woodrow.typepad.com/the_ponderings_of_woodrow/2006/08/the_word_enterp.html#comment-21287839
Yes, the beauty of empowering people is indisputable, that’s what motivates me in what I do. More accurate comment would be – “the danger of empowering people who want control but have no other outlet of power than open systems”.
Perhaps I am over-reacting and I am sure that there are many equally heated (and more) debates about entries all over the Wikipedia.
signed The outraged of Tunbridge Wells 2.0 :)
Adriana, I should begin by establishing my own political bias for a system of governance that is based on checks and balances. Having said that I recognize the fact that the term “author” also serves as a root for the term “authority.” I have no trouble accepting the premise that ever act of authorship is also an assertion of authority. As an aside, this puts me in opposition to Roland Barthes, who, in his “Death of the Author” essay proposed that the term “scriptor” be introduced as a replacement for “author” to avoid the linguistic connection:
http://en.wikipedia.org/wiki/Death_of_the_author
Note, however, that I used the word “premise,” rather than “fact!” The point of the two blog entries I cited in Comment #1 is that an EDITOR also has the power to assert authority. Having worked as both author and editor, I believe strongly that the relationship between these two agents is very much one of check-and-balance. This is why I have proposed elsewhere that the introduction of editorial authority is a good way to deal with the need for governance in settings such as Wikipedia and even collections of open-source software. I do not hold it up as the only way, but it is the way that reflects the influence of the Constitution of the United States of America on my thinking!
I need to think about that, my gut feel is that we agree but it’s a question of scale. One editor, a few editors, problems. Many editors, less of a problem. I would rather think about people being able to acquire editing rights simply, to be able to retain them simply.
In the context of your author discussion, what about voter? Voters express opinions because they are franchised to. Blogs allow people to vote with limited anchoring-or-framing. And ratings are like votes.
Just offhand thoughts.
JP, in that blog entry about the experiment at the German Wikipedia that I cited above, I compared the experiment with Xerox’ experiences with Eureka. This was sort of a pre-blog through which repair technicians could share experieces (“war stories”) about particularly difficult problems at the customer site, where the “approved” documentation was, at best, useless. These experiences were written out in free text by the technician and then passed to a validator, who basically acted as an knowledgeable editor. The primary function of the validator was to assess the soundness and generality of the author’s experience, but there was also a secondary function of making sure that the author did a good job in communicating that experience to the rest of the community of repair technicians. So, this was a one-author-one-editor situation that worked very well; and, as in indicator of the AUTHORITY of the content, when the text was made public, the names of BOTH the technician and the validator appeared at the top!
My personal view of the technician-validator relationship is that it involves establishing CONSENSUS. This is not always easy, but it is usually manageable when only two people are involved. When you bring additional people into the mix, you run the risk of their being “too many chefs.” Consensus becomes more difficult to attain, and a failure to attain consensus may only be resolved by voting. Unfortunately, the more people involved, the more likely that votes will be based on socio-political matters, rather than the actual issue being deliberated. Le Bon was well aware of voting when he wrote THE CROWD, which I invoked in our discussion of Gresham’s Law; and I believe his cautionary observations are just as relevant to this convervation.
Probably the best way to summarize this position is by observing how frequently it is the case that a decision by vote has more to do with popularity than with substance. I doubt that any of us would like to see such a decision strategy determine the content of Wikipedia. I also suspect that a Xerox repair technician with a REALLY hard problem would like any advice he receives to be informed by something more warranted than popularity!
Consensus has always intrigued, sometimes even puzzled me. I think it boils down to trust and covenant relationships. Trust is not ephemeral. Consensus is easy when you have people who are committed to each other in a relationship; modern firms have attrition rates that make consensus harder, and as a result you land up with popularity based decision making, spin doctors, anchoring and framing, memes taking moral high grounds, that sort of thing. Are you familiar with Lakoff?
JP, now it is my turn to introduce a name into the pool! Are you familiar with the work of Karen Tracy? If not, a good way to start would probably be to check out the blog entry I wrote about her at :
http://blog.360.yahoo.com/blog-Mff23hgidqmHGqbcv.lfskakEtS6qLVHUEMFUG4-?cq=1&bid=58&yy=2006&mm=7&p=58
What is particularly relevant to the context of the current discussion is the distinction she draws between DELIBERATION and DECISION-MAKING in group meetings. A simple way to put it is that, if all else fails, a group can arrive at a decision through a vote; deliberation is much more subtle, since it is ultimately concerned with arriving at a consensus. In the American judicial system the verdict of the jury for certain serious crimes must be unanimous, and those are the situations in which it is accurate to say that the jury DELIBERATES. How they do this is, indeed, can often be puzzling and even intriguing. For Reginald Rose, who wrote TWELVE ANGRY MEN, deliberation was even dramatic; and this may get at some of your puzzlement. One of the ways to work through this puzzlement is to acknowledge Kenneth Burke’s distinction between “scientistic” and “dramatistic” reasoning. Decision support systems (and, for that matter, decision theory itself) are pretty much scientistic, because science is what those guys know how to do. Deliberation, on the other hand, is more dramatistic, being driven by “softer human” concepts like motive. If I ever finally enjoy the luxury of calling the shots for an educational system, I want to make sure that they is a proper balance of the scientistic and the dramatistic, if not a Hegelian commitment to a synthesis of these two moments!
I have been following George Lakoff since his metaphor work with Mark Johnson. I met him (and talked with him) at (at least) one cognitive science gathering. I am not as strong a camp follower as I used to be, though. To invoke Hegelian language again, Lakoff is very much conceptual-analytic, while my own tastes tend to run to the historical-genetic (which may be another sign of my dramatistic inclinations).