On 18 April 2012 14:13, Andreas Kolbe <jayen...@gmail.com> wrote:

> Well, it is still true that in a conventional encyclopedia, everything goes
> through vigorous professional fact checking *before* publication. We have
> nothing to compare to that. Not even Pending Changes. Surely that is a
> very, very significant difference indeed?

The attempt to equate Wikipedia with a traditional encyclopedia breaks down
in several places.

> As a result, the kinds of inaccuracies we have in Wikipedia can be in a
> whole different league than the sort of error you might find in Britannica;
> there is often active malice at work, as opposed to the occasional cock-up,
> and you are talking about the no. 1 Google link for a person or company,
> rather than something appearing on page 582 of a dusty tome that few people
> own, let alone read.
If the context is "media" rather than "pedia", you have to compare WP with
what newspapers do. And we know the arguments: newspapers will publish
corrections; they are reluctant to do so even with glaring factual mistakes
(which are very frequent); they almost never publish corrections with the
same prominence as the original; and there is an argument that modifying
the original article online would be "wrong" (per the New York Times). It
makes it clear that the "journal of record" model has deficiencies also:
those in that business are not interested in annotation of errors.

Those who work around here are generally aware of WP's deficiencies. I
think pre-moderation (per the EB) isn't what we need, but extending
existing techniques to BLPs in a broader way probably is. This thread has
been fairly effectively diverted by Tom D's aside. In the broader context,
there is no competing "media" model that serves the "pedia" function in an
obviously better way. Given the range of options available to us,
"lo-bandwith pedia" is probably a worse choice than salami-slicing the BLP

(One day we may decide enWP needs a separate community to patrol BLPs,
effectively forking the wiki. I wouldn't rule this out, say when we hit 5
million articles, but the results might not be  so very different. I wonder
if Wikidata would be able to underpin a better model, in time.)

WikiEN-l mailing list
To unsubscribe from this mailing list, visit:

Reply via email to