[WikiEN-l] "How the Professor Who Fooled Wikipedia Got Caught by Reddit", _The Atlantic_
Anthony
wikimail at inbox.org
Wed May 23 19:20:39 UTC 2012
On Wed, May 23, 2012 at 2:23 PM, David Levy <lifeisunfair at gmail.com> wrote:
> Anthony wrote:
>
>> > > What established framework are you talking about, here?
>
>> > I'm referring to Wikipedia's policies and guidelines (and more
>> > importantly, the underlying principles).
>> >
>> > An editor, acting in good faith, might believe that creating pages
>> > for dictionary definitions or dessert recipes improves the
>> > encyclopedia. Does this mean that we're required to refrain from
>> > intervening? Of course not.
>
>> Of course not. You should revert the editor's changes.
>
> Exactly.
>
> You stated that "trusting people to act in good faith in the way that
> they feel is in the long-term best interest of creating an
> encyclopedia is what Wikipedia is all about". My point is that
> additional criteria are routinely applied. Someone's good-faith
> belief that a particular act "is in the long-term best interest of
> creating an encyclopedia" doesn't automatically justify (let alone
> mandate) its acceptance by the community.
You certainly should revert Gwern's changes. There's no dispute about that.
>> The data may still be useful.
>
> Agreed. I don't assert that the experiment is invalid. I note that
> *others* do.
Which others? I thought you were referring to me as one of the others.
>> > Maybe the community cares.
>
>> Then the community can come up with its own experiment. Or, they can
>> if you'll let them.
>
> If the community devises a consensus-backed experiment, of course I'll
> "let them".
Heh. What's a "consensus-backed experiment"?
More information about the WikiEN-l
mailing list