[Wikipedia-l] Wikipedia moderators and moral authority

Mon Nov 11 21:10:00 UTC 2002

Hello Larry!

>>> It would be nice to get some specific examples. <<

> How many specific examples would it take to convince you, I wonder?  I
> doubt I could produce enough, because our difference is philosophical.

I don't think so. I respect the current system of sysops. I think it can  
be improved, but I'm not against authority. Unlike what you may have  
assumed at this point (I get the feeling you read the mail while you  
replied), I'm not an anarchist. I believe in direct democracy and  
therefore in the necessity of enforcement.

The main problem I have is that right now, we are not a democracy. You may  
call it mob rule, an elite, whatever -- decisions are generally made by  
individuals, though, and not by a collective. I want to create more  
informed decisions by involving more people in a well-defined voting  
process with a discussion phase and a decision phase (where these phases  
could be very short in emergencies -- user definable).

While I believe such a process could be quite streamlined, I think it  
should generally be used only when other, "softer" means don't work. This  
is certainly the case with all vandals: here we have no choice but to  
intervene. I am more lenient (and that is indeed a difference of  
philosophy) towards "cranks" and the like.

By asking you whether the problem you perceive is growing or not I did not  
mean to imply that there is no problem. The question whether it is  
growing, however, is important to determine the urgency of the solution.  
How much time do we have to implement something meaningful?

Working on Wikipedia today, I agree more with you than I did yesterday.  
About half of all the anonymous edits were vandalism. It is my  
*perception* that this is significantly more than when I started. I'd  
still like to quantify this, though. It's quite possible that we are now  
primarily getting new users via Google, and the percentage of those who  
are valuable contributors is possibly smaller than with our earlier  
contributors, with a higher number of vandals.

>>> Perhaps Ed's "Annoying users" page (if renamed) isn't such a bad idea. <<

> Hysteria?  I have to support Zoe here; Lir is a disruptive child, and she
> should probably be banned.

I disagree here, but I'm not familiar with all evidence (which is why a  
"Problematic users" page, as a record of evidence, might be helpful). I  
followed much of Lir's actions today and saw nothing too problematic.  
She's antagonistic, sometimes silly and certainly not as smart as she  
thinks, but I believe we can deal with here as long as she doesn't  
vandalize pages.

The "hysteria" remark referred to a specific incident and was probably an  
overstatement, I apologized to Zoe on her user page.

> No.  Our collective experience is more than enough proof that we need
> *consistent* enforcement, Erik, not *stronger* enforcement.  Right now we
> have very strong enforcement.

I don't think so, IP banning is quite ineffective, we clearly need to work  
on this. Unlike Jimbo, I'm not so sure we're dealing with the same vandal.

But I also agree that enforcement needs to be applied consistently. I'm  
worried that we may have missed vandalism when sysops weren't available.  
But (more below) I don't think that the problem really comes from our  
logged in userbase.

>  Your interest is obviously not in reducing
> power but in keeping power distributed among a lot of different people who
> use it in totally different, inconsistent ways, and none of which has any
> particular respect among other users.

Actually, no, I believe we need to quantify the generally agreed upon  
standards using open voting. I have described such a system already on  
wikitech, and have explained why I believe voting is important here on  
wikipedia-l.

I am for enforcement in what you rightly call "relatively extreme" cases.

> If we had power concentrated in the power of a rotating group of trusted
> individuals

There are many problems of this idea vs. open voting, such as definition  
of "trust", randomness of assignment (Slashdot-style voting) etc., I  
pointed these out already.

>>> That's not the kind of list I'm talking about, because it only tells us
> about the reactions, not the actual actions. You may say that these people
> were driven away by silly eedjots, but I cannot tell whether this is true
> without looking at the actual conflicts. <<

> Be serious--look at what you just wrote.  Does anyone other than you
> really need it to be proven?

Perhaps, perhaps not. You're talking to me now, though, so either present  
to me the knowledge I do not have, or you cannot presume that I have it.

>>> Often I've seen so-called experts on Wikipedia try to stop reasonable
> debate by simple assertion of their authority. This doesn't work, and this
> shouldn't work on Wikipedia, and if they can't handle that fact, it's my
> turn to say they should better leave. <<

> Wait a second.  Take a step back and put this exchange into context.  I
> said that Wikipedia is descending into a sort of mob rule, and that this
> has driven away, and will continue to drive away, some of our best
> contributors.  You reply, here, by defending the mob against the experts.

No, that's a grossly incorrect characterization of my statement. You see,  
I asked my question above for examples of the conflicts because I myself  
have had conflicts with so-called experts, and I did not accept their  
assertion of authority as an argument, which offended them.

Now I would hardly call myself part of a "mob", and I would find it  
demeaning and insulting to be called so. I can draw primarily from my own  
experience when it comes to conflicts, so what I can say is this: If this  
is the kind of conflicts you are talking about, where you think the so- 
called experts should be able to enforce their POV just because of their  
status as experts, then we are clearly of extremely opposite opinions. I  
do not support an expertocracy, and I don't think you do either. In the  
above paragaph, it sounds like it, though.

>> I'm talking about a body of trusted members, not an "elite."  Tarring
>> the proposal with that word isn't an argument.

>>> Sorry, but I do not really see much of a difference. A group with
> superior powers is an elite, trusted or not. The word "elite" requires a
> certain stability of that position, though, so it might not apply to an
> approach of random moderation privileges. <<

> That shows that you're essentially viewing this ideologically, and that
> you're expecting the rest of us to buy an essentially anarchistic
> ideology: *any* group of trusted members who has powers others don't have
> is *by definition* an "elite."  But in the mouths of any libertarian or
> anarchist, "elites" (and "cabals") are necessarily evil.

I'm not a libertarian or anarchist, and I do not consider elites  
necessarily evil (I do not believe in a concept of evil). I just don't  
think they are necessary either.

> The result of this ideology is..

Please, don't assume so much.

> They, like every committed Wikipedian, can't help but derive pleasure from
> the fact that we have built up a huge structure of knowledge.  But they
> have a woefully incomplete idea what makes it possible.  What makes it
> possible is precisely the *combination* of freedom, which makes it easy to
> contribute, *and* enforced standards, which define and guide our mission.

As I said repeatedly, I believe in enforcement, but only as a last resort,  
and only democratically justified. I also find enforcement more important  
in cases of vandalism than in cases of "cranks".

> to a project defined by rules, with a particular purpose.

>> Well, the times I'm concerned about aren't necessarily times when people
>> are shouting against the majority, but when they write nonsense, brazen
>> political propoganda, crankish unsupported stuff, and so forth--in other
>> words, violating community standards.

>>> Do you mean nonsense in the sense of "something that just isn't true"
> or in the sense of simple noise, like crapflooders? How do you plan to
> define / recognize "crankish unspported stuff"? <<

> Do you really think it would resolve anything in our discussion if I were
> to supply you with an answer to these questions?  No, you seem to want to
> ask rhetorical questions, and the point of the questions is: there are no
> clear standards whereby we can determine when community standards are
> violated.

That's not true. Ed has proposed some good standards, although I agree  
with Cunctator that it is not always easy to identify partisanship (but  
there are egregious cases). Our policy should be to argue with people, to  
revert changes a few times to help them understand the wiki principle, and  
if they don't agree with our minimum community standards, we enforce them  
by community vote.

I also think, however, that your phrases above are much more vague, and  
much less apt for inclusion in a set of minimum standards.

> OK, so you're worried that "crankish unsupported stuff" and other words
> we'd use to describe undesirable material and behavior would be such that
> we'd disagree about cases.  Of course we would.  But if we're reasonable
> people and understand what an encyclopedia *generally* requires, and we
> have much experience actually working on an encyclopedia, then we can
> certainly agree on a lot of cases.

Hopefully so, but because of the potential for disagreement, I think the  
decision making process should remain open.

> The lack of absolute unanimity in every case does not--just to give an
> example--provide people an excuse to write unsupportable or provably false
> stuff in articles, just for one example.

True. However, it's perfectly OK to attribute provably false stuff. Ed  
Poor does this all the time when he writes about homosexuality :-)

> It doesn't mean we can't
> forthrightly eject (or completely rewrite) material that is, on any
> reasonable person's view, a violation of our neutrality policy.

I agree.

>>> The idea of soft security has evolved in wikis, and it is only fair to
> point you to the respective page at MeatballWiki for the social and
> technical components of soft security:
> http://www.usemod.com/cgi-bin/mb.pl?SoftSecurity <<

> I'm not particularly interested in going to the website to find out what
> you mean.  If you want to introduce an unfamiliar term into a debate, it
> is polite to define it.

It is also polite to visit the links provided in someone else's mail in a  
discussion. Are you afraid of MeatballWiki because you associate it with  
TheCunctator? Rest assured, he's not much more popular around there than  
he is here. MeatballWiki is a pretty good resource for many questions we  
are debating, much more rich in content than the meta-wikipedia, and there  
are many smart people over there. SoftSecurity as defined by the Meatball  
people involves a large amount of different components, usually social  
concepts relying on existing wiki technology as opposed to extending it.
We have talked about some aspects already. I suggest you familiarize  
yourself with the concept.

> >> We already *have* a
> situation where we occasionally ban legitimate users and delete legitimate
> material. We need to get away from this. I am absolutely disgusted by the
> thought that we are already banning completely innocent users. <<

> Erik, two things.  First, we already have banning and deletion.  We aren't
> debating about those, but what you write above makes it sound as if we
> were.

No, no, no. I'm not against banning or deleting. I'm for making the  
process less error-prone by involving more people in it. We are banning  
innocent people behind multi-user proxies, and sometimes overzealous  
sysops delete allegedly infringing material which is completely harmless.  
This is, in my opinion, a very serious issue that needs to be addresed.  
You believe that your trusted moderators could make better decisions, and  
it's quite possible that in clear-cut cases, they would be better than  
what we have now (though in less clear-cut cases, they might be worse). I  
think open voting would be the best solution.

> Second, the mere *existence* of sanctions hardly implies that those
> sanctions will be abused to any degree at all.  If your point is simply
> "Power can be abused," I'm totally unconvinced and I don't think anyone
> else will be either.  Surely you must have more reason to support it than
> that.

It's quite possible that abuse will be minimal in a scheme like yours. I'm  
afraid, however, if broad policies like "crankish unsupported stuff" are  
adopted, a high degree of arbitrariness is introduced. We need to be very  
precise here to avoid the Everything2/Slashdot dilemma.

> After all, as for example under my rough proposal, we can have strong and
> multiple safeguards against abuse.

I like that about your proposal. It makes it closer to mine :-)

>>> My underlying philosophy here is that it's worse to punish an innocent
> man than to let a guilty man go free. <<

> But that's hardly a reason never to punish anyone for anything, is it?

No, it is not. Repeat violators must be "punished" as a matter of self- 
defense.

>>> I think this kind of last resort authority should not be concentrated
> but distributed. If a poll shows that many members think that member X has
> "crossed the line", this sends a much stronger message than any non-
> totalitarian scheme of concentrated authority. Especially if last resort
> measures like banning *can* be approved by the majority. <<

> You favor a democracy, susceptible to mob rule as at present

Please explain "suceptible to mob rule as at present". Currently we have a  
situation where decisions are made by individual sysops, they are not  
verified by anyone else. In my proposal, they would be verified by anyone  
interested in doing so (logged in users, possibly only users with >n  
contributions to avoid vote flooding).

> I favor a republic,

It was my understanding that you are libertarian. How is that compatible  
with that view? (Yes, I know you're only using a metaphor -- but is real  
life group decision making that much different from Wikipedia?)

>>> Yes, I agree that those cases exist. But I also believe that we need to
> have a lot of patience when dealing with newbies. Not an infinite amount,
> but a lot. And I think everyone should be given the opportunity to
> rehabilitate themselves. <<

> Well, I do agree with that, and it's not at all inconsistent with my
> proposal.

I think most of the supposed inconsistencies are simply misunderstandings,  
but the main difference in opinion is that you regard cranks as the most  
serious problems and I (and probably most sysops) see vandals as much more  
annoying and serious.

> I think our experience clearly indicates that we can't make people nice if
> they don't want to be.

Well, no offense, but I don't think you're particularly good at making  
people nice, so that may have something to do with the experience :-)

> Anyway, it's the nearly-worthless contributors who
> on balance damage the project by the dross they shovel in (and the
> resulting controversies and wasted time) that are the real problem.

I disagree here. I don't see a high number of "nearly-worthless  
contributors" (that you characterize human beings as worthless is another  
story). You named Lir (even there I don't agree entirely), who else?

>>> Well, I can predict that I'm going to "attack" experts myself if they
> add non-NPOV content, fail to cite sources properly, insist on their
> authority to make their point etc. <<

> Good luck.  Remember, experts know more about their areas than you do.
> That's why we call them experts.

That's incorrect, they're *supposed* to know more about their areas than  
non-experts. Often they don't. Sometimes they're even paid to lie,  
especially when commercial interests are involved.

> I'm a Ph.D. epistemologist.  My dissertation adviser was (still is) an
> expert on the concept of expertise, and I've read several papers on this
> area of social epistemology as part of a graduate course.  In addition, I
> gave careful thought to this subject while working on Nupedia.  Now, what
> is that you think my view of what experts are and what makes an expert?

I haven't read your adviser's papers, I base my opinions solely on your  
mails. It appears that you value credentials such as a degree or  
publications very highly, as opposed to reputation building through  
verifiable expertise, regardless of background. An expert, in my view, can  
be pseudonymous, he can be a 13 year old kid, it can be a motivated  
housewife. A degree and other credentials are only one way of measuring  
reputation, and not a particularly good one, because many institutions are  
highly biased. We had two certified mediavalists on Wikipedia, who both  
added highly biased POV material from a Christian-apologetic perspective,  
for example. You yourself added quite a bit of NPOV material (often, but  
not always, marked as such).

> I just have no idea why you say the system I proposed would "lead to less
> informed decisions."  Perhaps you should reread the proposal (even though
> it was, as I said, just a rough outline); I even went so far as to suggest
> that perhaps there would be a body of Wikipedia "case law" developed, that
> moderators could consult.  This would lead to *less* informed decisions
> than in the present case, when virtually anyone can have the power to ban
> and delete?

Yes, in some cases, because no longer the people who make the decisions  
are the ones who care about the matter, but instead the ones who were  
assigned moderation duty by your random number generator. A voting system  
I propose would be similar to the Recent_Changes page, you would look at  
the polls you find interesting, read the arguments, then vote. In your  
system, however, the people who vote/make the decision are simply  
assigned, and they may not care at all about the debate at hand and just  
throw in a quick "yes or no". Because of your safeguards, this is better  
than an individual sysop in the clear-cut cases, but possibly worse in  
less clear-cut cases. And I consider this especially problematic because  
you seem to see the less clear-cut cases as the more serious problem.

Regards,

Erik