There is no general solution to this problem, by way of background:
http://tech.mit.edu/V123/N8/8voting.8n.html
http://www.investopedia.com/terms/a/arrows-impossibility-theorem.asp
http://en.wikipedia.org/wiki/Arrow%27s_impossibility_theorem
http://cowles.econ.yale.edu/~gean/art/p1116.pdf
There is no general solution that allows all preferences in all orders, selects all group
preferences in order, and forbids dictators – that is individuals that break ties.
Wikipedia as with anywhere else. What can happen is that layers can be put in place, and a
priori choices made. Wikipedia picked non-dictatorship very highly. The structure of
wikipedia is meant to reduce the roll of dictators, and particularly ones that have a
stake in the matter. It is imperfect at this, and two common problems are paid editors,
and people who want to violate transitivity. On wikipedia, dictatorship is delayed as much
as possible, and it is unlinked to some extent from preference.
While wikipedia is very direct democracy most of the time. Academia has chosen feudalism.
PIs (Primary Investigators) are dictators in their own area, and owe fealty to some larger
institution. They accept a system of mediation, which includes peer review, as part of the
grant of power. This system is wildly imperfect, and going through convulsions,
convulsions brought on by the same realities that created Wikipedia: the high cost of
gatekeepers and the reduction in the value they add. In Wikipedia's case, standard
encyclopedias don't spend enough on writing, and a great deal on maintaining their
position, the case of academia, the high cost of academic journals, who spend all of their
money on staying important – articles are written for free, and peer reviewers work for
free.
There are forums for gather pre-consensus, including talks, conferences, starting ones own
journal, open source journals, and ArchiveX. Wikipedia is not one of these. It is not a
community for the vetting of utterly new ideas. It can alter consensus, because often the
public discourse is dominated by money and network effects, which involve themselves in
suppression of other ideas, that is violating transitivity and admissibility. By balanced
and neutral coverage, it gives general intelligent readers access to a source of
information, which is directed as explanation. Read the links. One is a short bastardized
example, for investopedia, one is a kind of coffee hour discourse, from hit, one a
moderately technical exercise without enough context. I will submit that the best
introduction to the theorem, and its context, is wikipedia, because it is not meant to
advance a specific career or to be narrowly tailored to a small audience. The system as it
is can be painful, but it does work better that various forms of dictatorship or limited
admissibility. The purpose of Wikipedia is to move information from early adopters to the
early majority and from there to the late majority. It is not for creating early adopters,
and is, in fact, intended to be moderately hostile to people trying to do this.
Experts can, and should upgrade articles to include accepted though not preferred ideas in
their field, move from basic to advanced content, and clarify explanations. But to expect
that editing article A in the standard form gives you the permission to insert original
research in B is to violate transitivity.
On Oct 27, 2012, at 4:41 AM, David Gerard wrote:
On 26 October 2012 23:49, Apostolis Xekoukoulotakis
<xekoukou(a)gmail.com> wrote:
Progress only occurs when there are different
views that can coexist. When
you force the less popular views to disapear into a talk page, ie into
obscurity, then the stronger view becomes the only view.
So wikipedia is not solving the filter buble problem as it is now. It
simply has one filter buble created by the most enduring, by the strongest.
In my proposal, all different opinions could and should be easiliy
accessible.
POV forked articles are a perennial proposal. They have in fact been
implemented on other wikis, such as WikInfo. However, it seems in
practice that readers want one place to go, not several.
- d.
_______________________________________________
Wikipedia-l mailing list
Wikipedia-l(a)lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikipedia-l