On the advice of one wikipedia admin I've already recruited, I'm
emailing this list to seek a few others to help me with a research
experiment I'm hoping to undertake. Basically, I'm interested in
trying to graft Wikipedia's highly effective consensus-editing model
onto some currently-jammed political discourse.
I've become interested in (the lack of) productive political discourse
on the web---the polarization and flame wars that seem to be the
norm---and I'm beginning to think about ways to improve the situation.
In particular, I'm thinking about social tools that can encourage and
help groups to understand their differences and reach agreement _at
least_ on the underlying facts of an issue even if they draw different
conclusions.
A particular example that I find quite interesting is on Wikipedia.
There is an article about the Obama health care plan and, more
interesting, an article about the debate about the plan. The latter
allows users to "go meta" and present some highly non-NPOV opinions
about the plan by writing neutrally about the fact that some people hold
those opinions. This seems like a really good model for how people who
strongly disagree could nonetheless work in good faith to map out the
issues of the disagreement.
Recently I met the founder of
politifact---http://politifact.com/ ---a
cool site that "fact checks" statements by politicians. Apparently,
each time they publish one, they get a slew of angry responses on
twitter/facebook about how wrong they are---from _both sides_ of the
discussion.
Since this is a clear case of an underlying _fact_ that is being
checked, it seems like an obvious target for reaching consensus.
So we brainstormed an experiment. We recruit those angry folk who think
politifact got it wrong, and see if they can work together to figure out
the "right" answer. Ultimately, we imagine creating some tools to help
do this. But we wanted to start with some experiments to understand how
the discussion process might play out. So we though to start simply by
locking all participants in a room---ie, a wiki page---and letting them
hash things out there. Obviously that's going to require some ground
rules---which we can lift straight out of Wikipedia, taking the page
about debate about Obama's plan as a model. But equally obviously,
it's going to require some _enforcement_ of those ground rules. Since
Wikipedia admins are familiar with the ground rules, the enforcement
process, and the tools supporting it, it seemed natural to see if we
could recruit some of you as the policemen to walk that beat.
So that's the story. If you're interested in participating, please
contact me.
thanks
--
David Karger
Professor, EECS
MIT Computer Science and Artificial Intelligence Lab
32 Vassar St.
Cambridge, MA 02138
http://people.csail.mit.edu/karger