At 06:41 PM 1/27/2010, Jed Rothwell wrote:
Abd ul-Rahman Lomax wrote:
It's really an aspect of the problem of scale. Those who could do
something about it are overwhelmed and must make snap judgments, so
when an issue is complex, really bad decisions are made.
This is true, and it is difficult problem. Sometimes, this is what
causes capable people in the top ranks of huge organizations to make
horrendous errors. For example, in the Federal Gov't, or at IBM or
GM. It seems likely to me that Obama or the head of the DoE have no
knowledge of cold fusion, for example, because they have so much
else on their plates, and so many people giving them advice. They
have no time to hear about cold fusion. No one in their office
happened to see "60 Minutes" last April. (I suppose . . .)
What I've been suggesting is to understand the mechanisms by which a
general consensus is overthrow, the ways in which fringe ideas that
have an actual basis can (and do, eventually) gain wider consideration.
Instead of going for Obama, find who has Obama's ear and who might be
willing to take the time to understand the topic. And if you can't
find any such person with the time (good chance), then someone who
has the ear of the one who has the ear.
And then another, so that it comes in from two different sources.
When several people start mentioning Cold fusion to people close to
Obama, the message starts to punch through the noise.
This also explains why skilled generals in the heat of battle
sometimes make huge mistakes that are out of character. The press of
events, fatigue, or the need to make snap decisions without enough
information causes them to make mistakes they would not normally make.
Right. Hence a truly skilled general surrounds himself with people
who criticize his proposals. By nature, the office of general is one
where a decision must be made, but to fool a well-advised general is
much more difficult than to fool one who only surrounds himself with
sycophants.
You have to sympathize with the Wikipedia Foundation in this regard.
When a method generally works but occasionally causes disastrous
failures it is hard to say they should abandon it.
That's right. And, in fact, they should not abandon the method. They
should modify it with structure that detects the errors and escalates
efficiently when it's needed. They also need to stop requiring
Sisyphus to roll the boulder up the hill over and over, and the
software tools exist for what's called Flagged Revisions. But Flagged
Revisions requires a set of editors trusted to be able to set the
flags, and the community has become paralyzed, unable to make
decisions on a large scale. And there is no mechanism for doing it,
in fact, because the whole of Wikipedia operates as an adhocracy or
ochlocracy, avoiding the making of actual deliberated collective
decisions. It can be fixed, but the conservative forces on Wikipedia,
clinging fervently to the status quo, are formidable.
The free-for-all technique does not work for an article on cold
fusion, but it works for hundreds of thousands of other articles,
and many of these would not even be written in the first place with
a tighter set of rules. Articles about Japanese comic book
characters, for example, would not be written. They have some
social and literary value for people who want to learn about Japan.
See, for example:
http://en.wikipedia.org/wiki/Maison_Ikkoku
Not important, you say? Maybe not, but neither is most literature.
It is a good way to learn about what it was like living as a college
student in Japan in the 1980s.
If I didn't think Wikipedia was important, I'd not have devoted
several years to it.... I do believe I know how to fix it, which
doesn't translate to "instant fix."
I just suggested on the major Wikipedia mailing list a solution to
what has become a huge flap over unsourced biographies of living
people. A bot was developed to find these articles and automatically
delete them. Bad idea, actually, but there is a good idea which is
very close to it! There may be something like 80,000 of these
biographies, with more being created all the time.
The idea isn't a new one, it's called Pure Wiki Deletion, which
refers to blanking content rather than actually deleting it.
Strictly, with these, the content would not be blanking, it would
instead be redirected to a page which explains the problem with the
article, and which then provides instructions to how to read what was
there, and to restore the article. A bot could do this in a flash, it
fixes the legal problem with the articles immediately, it leaves the
content where anyone can read it, warned about the unreliability, and
anyone can fix it, and, then, activity "fixing" these articles can be
monitored. Note that "actual deletion" isn't really the case with
Wikipedia, content is not deleted, it is, rather, hidden from all but
those with administrative privileges. There is true deletion, called
"oversight," but that is normally limited to illegal content of some
kind, and ordinary administrators can't do it. (In fact, I think that
even oversight isn't complete deletion, which takes the highest level
of privilege, "developer." But I'm not sure about the details.)
Another editor also suggested the same thing. The only response was
from one very "reputable" editor whose comment was something like
"Don't you think that something new shouldn't be tried for the first
time on the largest wiki in the world?" But, in fact, PWD is called
"pure wiki deletion" because it is already in use, just not with the
exact application, not with blanking whole articles. It is how most
violations of content policies are already dealt with, by reverting
the edits. Anyone can see them and restore them, normally.
When an article is deleted and there is a link to it, the link is
called a "redlink" because it displays that way. This, then, in
wikitheory, invites people to create the article. If the article is
deleted, the person following a redlink may get a message that the
article was deleted, and suggests caution before recreating it, plus
recreation is extra effort, effort that would be better put into
finding sources for what was already there. Hence deletion is very
inefficient, compared to stubbing (taking an article down to the bare
minimum that can be reliably stated from sources) or blanking with an
explanation.
So, very simple solution that satisfies, in fact, all the arguments
being presented on the two sides of the on-wiki debate. It
immediately gets rid of the unsourced article problem. It leaves
inviting information in place. It requires no extra effort to review
and maintain, errors in PWD can be found and corrected by anyone. (A
category tag would be added to all articles blanked by bot, possibly
to the Talk page, so anyone could review these.) And what happens?
It's not even noticed. I was even brief, which takes a lot of effort.
What I conclude is that they would rather fight each other than
actually solve the problems. There is no structure to filter and vet
and polish ideas as they are escalated to wider attention. There is
no structure for decision-making that doesn't radically break down if
more than a handful of editors participate in it. In a word, there is
no deliberative process, it is quite like a mass meeting with nobody
understanding how to do it, and anyone who tries to point out that
how to do it was worked out centuries ago and it still works (and can
work even better with the internet) is an outsider, obviously, and
should be banned for meddling. And, believe me, that happens all the time.
And, yes, this can be overcome. I predict. Maybe I'm wrong, and maybe not.
Perhaps we need one set of procedures for comic-book fans and
another for physics. After all, in the real world information on
these topics is written, reviewed and published by very different
means, with utterly different standards.
Yes. The guidelines are flexible enough to handle this, but the lack
of structure makes application horribly uneven.
There does not seem to be any acrimonious disputes in the talk
section of this comic book article, by the way. The article content
seems accurate to me.
That is more or less the norm for non-controversial topics. With
comic book characters, unless deletionist editors get wind of it and,
for example, apply the same standard as might be appropriate for a
science article to an article about "fancruft," the process works.
