On Tue, Jan 26, 2016 at 9:36 AM, Magnus Manske <magnusman...@googlemail.com>
> Be careful with that "obvious" word...
Things have been busy of late, and I never had time to properly respond to
this blog post of yours. (For anyone else who has forgotten, this was the
discussion about vast swathes of Wikidata lacking reliable references, as
discussed in .)
You say, "the impression I get from Andreas’ text is that, while Wikipedia
has some issues, references are basically OK, whereas they are essentially
non-existent in Wikidata." In your piece, you then go on to compare the
referencing density of Wikidata content to that of Wikipedia content,
finding that Wikidata, even now, doesn't do at all badly compared to
You present it as a sort of sibling rivalry: if Wikipedia doesn't do any
better herself, why does she complain about her sister Wikidata? I recall
Denny and Gerard making similar arguments.
In doing so, you miss the core point of the criticism. My point is that
Wikipedia's *referencing standards are okay*, and that *those* are what
Wikidata should be assessed against.
Wikidata and Wikipedia have very different purposes: Wikipedia is an
encyclopedia to be read; Wikidata is a database. No one reads a database.
The whole purpose of a database is to have its content multiplied and
surfaced elsewhere. Therefore it is even more essential that its content
stand on solid ground.
If you want to measure Wikidata against something else, you should measure
it against the sources that open knowledge currently relies on, i.e. the
quality standards underlying WP:V, WP:RS and so on, especially if Wikidata
will also be used as a source in Wikipedias.
My argument has never been that Wikipedia content is good, and Wikidata
content is rubbish. The quality of Wikipedia's content is extremely
variable. Sometimes it's alarmingly unstable, and you see Wikipedia "truth"
shifting from one extreme to the other (example: ). Sometimes it's
manipulated (example: ). Wikipedia contains *a lot* of rubbish,
alongside some undeniably good content.
It's for that reason that I view it with dismay when Wikidata makes
wholesale imports "from Wikipedia", without so much as traceability to a
specific article and article revision, and a check whether the information
taken from Wikipedia was accurately sourced there.
At the Wikipedia Weekly Facebook group, we recently discussed the use of
Wikipedia as a source for legal decisions. On a human level, it's
perfectly normal and understandable for Wikimedians to feel validated, to
feel pride whenever a court makes such use of Wikipedia. But in my view,
one of the core tasks of the Wikimedia community should be to *discourage*
such use, and teach the legal profession Wikipedia literacy. This includes
at its most basic level not putting any faith into any statement in
Wikipedia *per se*, but instead checking and assessing its sourcing on each
and every occasion, and referencing the source instead. We all know that
complete nonsense can survive for a long time in Wikipedia, even in highly
> On Mon, Jan 25, 2016 at 1:56 PM Andreas Kolbe <jayen...@gmail.com> wrote:
> > On Fri, Jan 22, 2016 at 9:34 AM, Magnus Manske <
> > magnusman...@googlemail.com>
> > wrote:
> > > What you hear is "Wikidata is unreliable" (compared to the respective
> > > Wikipedia; proof, anyone? Please, show me proof; silence or anecdotes
> > don't
> > > count)
> > Any non-trivial content you want to add to Wikipedia today has to fulfil
> > one basic criterion: that the content be traceable to a professionally
> > published source.
> > Most Wikidata content fails that criterion. It's blooming obvious that
> > Wikidata is "unreliable" according to Wikipedia's definition of a
> > source", isn't it?
> >  https://tools.wmflabs.org/wikidata-todo/stats.php
> >  https://en.wikipedia.org/w/index.php?title=Wikipedia:SPS
> > _______________________________________________
> > Wikimedia-l mailing list, guidelines at:
> > https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
> > New messages to: Wikimediafirstname.lastname@example.org
> > Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l,
> > <mailto:wikimedia-l-requ...@lists.wikimedia.org?subject=unsubscribe>
> Wikimedia-l mailing list, guidelines at:
> New messages to: Wikimediaemail@example.com
> Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l,
Wikimedia-l mailing list, guidelines at:
New messages to: Wikimediafirstname.lastname@example.org