[WikiEN-l] NPOV and how to find and maintain it

[Abd ul-Rahman Lomax]

At 01:57 PM 5/10/2009, Sam Korn wrote:
>On Sun, May 10, 2009 at 3:34 PM, Abd ul-Rahman Lomax
>
>I meant that resolving the meatspace Israel-Palestine conflict is
>beyond the capabilities of the Committee.

Certainly. But if Wikipedia potential were realized, though being 
*better and more efficient* at following the core policies, it is not 
impossible that it would contribute to a real life resolution. Part 
(not all) of the RL problem is a lack of knowledge by each side of 
the reality of the other side and the reality of the historical 
situation. Knowledge helps real life.

Or else why is this whole project important? MMPORG?

> > The key to understanding this is, first of all, that NPOV isn't a
> > thing, a fixed state, a property of text in itself, it is a balance
> > that represents consensus.
>
>No.  NPOV is not determined by consensus.  Wikipedia's content is
>determined by consensus with NPOV being the guiding principle.
>Something does not become more neutral because fifteen Wikipedia
>editors say it's neutral.

Sam's opinion is common, and quite incorrect. What's missed in his 
understanding is how NPOV is determined, both as Wikipedia policy, 
but in real life. We did not invent NPOV.

Sure, NPOV is a guiding principle, but how do we recognize it? If we 
have a POV, we may be unable to recognize this POV as not neutral. If 
a POV is a mere majority POV, we may easily think that it's neutral, 
and the only way to recognize it is not is to pay attention to those 
who object. We can't always do that with every POV, we'd be doing 
nothing but examining the foundations of our beliefs, and we do need 
to take time for ordinary maintenance of the project, editing 
articles, and even eating and working and sleeping, though those are 
of lesser importance, right?

Something does not become more neutral because every single person on 
the planet signs on. However, if everyone does, it makes it pretty 
likely that it's neutral, don't you think? If everyone signs on, not 
only to the truth of the statement, but to a conclusion that the 
statement is neutral, that would pretty much seal it as well as it 
could ever be sealed, don't you agree?

> > We can measure NPOV by the percentage of editors who agree with a
> > text, and our goal should always be 100%.
>
>No.  The mere fact that no-one complains that their point of view is
>under-represented does not mean that it isn't.

Of course not. Especially if we've blocked everyone who dared. Sam 
isn't using what is called "charity of interpretation" in works on 
the philosophy of science, i.e., that claims are read 
sympathetically, to see what's right with them, not to imagine what 
is wrong and then object to the imagination. Suppose the "agreement" 
I posited is simply "agreement that the text is neutral and in 
conformance with policy." Would this change the response?

And, remember, I'm not proposing absolute truth, but only what we 
will sensibly rely upon as true. If nobody thinks that a text is POV, 
should we still treat it as such? What's the alternative?

What I'm proposing is an *objective* standard for NPOV that treats it 
as a goal to be approached, even if it is never reached. That is, 
instead of insisting that project text be neutral, which may be 
difficult or impossible in matters where there is any controversy, we 
insist that it approach neutrality as closely as possible, with 
"neutrality" being defined by the relative lack of objection. If we 
take steps to ensure that there is knowledgeable participation in 
this determination, we make it even more secure. We won't reach 100% 
in controversial matters for two reasons: the existence of unresolved 
but possibly legitimate controversy over the text, which may take 
time to complete (sometimes a lot of time), or the existence of 
editors who are firmly attached and unable to identify neutral text 
when they see it, they will not even consider it, since they may 
dislike obvious conclusions. (This often happens because the real 
dispute is elsewhere, not actually with the subject of the text 
itself, the text is merely a pawn in a different game.)

>It is no more possible to create neutrality by public vote than it is
>[[wikiality|to create reality by public vote]].

Did anyone suggest voting here? Note that consensus can *seem* to be 
like voting, but it is fundamentally different. We can know that 
consensus is not reached from a single dissenting comment. And if 
half the registered editors of Wikipedia were to vote, setting aside 
the damage to the servers, we would not know that we had reached 
absolute consensus. (Though it would be likely, that's a big sample!)

I wasn't suggesting *at all* that neutrality was created by !voting. 
Rather, polling is a tool that can be used to estimate the degree of 
consensus, sometimes, which can save a lot of work. Suppose that we 
have an editor arguing some piece of text. Other editors waste a lot 
of time debating with this person. Then someone gets the idea to ask 
if *anyone* supports the editor's position, other than the editor. 
Nobody responds. Until someone responds, don't you think that this 
"vote," in which no vote was cast, would effectively resolve the 
issue pending? And if we did this, we would be following standard 
deliberative procedure: no motion is debated unless seconded. 
Organizations routinely dispose of unsupported motions in less time 
than it takes to state the motion. The chair asks, "Is there a 
second," looks around the room, and then says, "Motion fails for lack 
of a second," which is a non-prejudicial close with maximal 
efficiency. Proposals may be *discussed* in "committee," or other 
informal session, or privately, but participation in that discussion 
is not binding or obligatory. So if Mr. Dedicated Fringe wants to 
discuss a proposed edit with anyone who consents, that's fine, but 
the lack of negative response does not indicate a consensus for the 
proposal, it indicates nothing except that, perhaps, N editors, or 
maybe a majority of those participating, which may be biased toward 
the Fringe contingent, support the proposal. When the edit is 
actually made to the article, and another editor who did not 
participate reverts it, we cannot claim that the revert is "against 
consensus," because a simple discussion doesn't establish consensus, 
nor does a single poll. True consensus is shown to be absent by the 
single revert! But, ultimately, we'd want to see at least one other 
editor agree with the revert, or, indeed, we may easily consider that 
it's "against consensus," depending on the level of support.

Consensus is found through *negotiation*, not by !voting. Polling is 
merely a tool that can sometimes make negotiation more efficient. In 
RL organizations with control over property, majority rule allows 
decisions to be made, short-term, quickly and efficiently, but if any 
organization cares about unity (and for ultimate survival, they 
better), they will not be content with mere majority for any decision 
of weight. For this reason, Robert's Rules requires a two-thirds 
majority to close debate on a motion. With more sophisticated 
deliberative systems, it's practical to go much higher than 
two-thirds and, in fact, to reach a level where debate isn't truly 
closed until there remain less than two editors willing to debate. 
(But actions are based on "rough consensus," we don't have to wait 
for full consensus to act, we merely consider that consensus may 
change, and we leave open paths for that to happen without 
disruption.) Wikipedia is actually there, it can be done now, but 
because we haven't recognized all this formally, it usually isn't 
done. The rest of the editors could ignore a debate, and say to the 
two editors involved, effectively, "Let us know if you come to an 
agreement, meanwhile, we have Other Stuff to do."