On 7 January 2018 at 02:05, Andy Mabbett <andy(a)pigsonthewing.org.uk> wrote:
Understandably you may not be aware that this is a fork of discussions
on Wikidata, where several of the points you raise have already been
addressed. See:
https://www.wikidata.org/wiki/Wikidata:Requests_for_del
etions#Artist_of_Black_Lunch_Table
and:
https://www.wikidata.org/wiki/Property_talk:P972#Abuse_
of_this_property_for_original_research
Thanks Andy, I had read those talk pages too as Erika helpfully linked them
in her first email in this thread, and I was still struggling to see the
argument behind the opinions presented.
Q28781198:Black Lunch Table is P31:instance of Q21025364:WikiProject
The objection is that anything entered into
P972:Catalog MUST be
P31:instance of Q2352616:catalog
Is that it?
No. While that is the case, it is also true that "catalogue (P972)" is
defined as "Either a) catalog where the item appears (used as
qualifier for catalog code (P528)), or b) catalog of an exhibition
(Q464980)"
No-one contends that 'b'; applies, and the absence of P528 values fails
'a'.
OK, thanks for highlighting that, I overlooked the strict significance of
the part in brackets
about use as a qualifier, and was led astray by the talk at
https://www.wikidata.org/wiki/Property_talk:P972#As_a_property which seemed
to support it being used legitimately as a property indicating membership
of a catalog without a catalog number.
Additionaly, this SPARQL query for 'valid' P972 properties
<https://query.wikidata.org/#SELECT%20%3Fitem%20%3FitemLabel%20%3Fvalue%20%3FvalueLabel%0AWHERE%0A%7B%0A%09%7B%0A%09%09SELECT%20DISTINCT%20%3Fitem%20%3Fvalue%20%7B%0A%09%09%09%7B%0A%09%09%09%09%3Fitem%20wdt%3AP972%20%3Fvalue%20.%0A%20%20%20%20%20%20%20%20%20%20%20%20%20%20%09%3Fvalue%20wdt%3AP31%2Fwdt%3AP279%2a%20%3Fclasses%0A%09%09%09%7D%20%20.%0A%09%09%09VALUES%20%3Fclasses%20%7B%20wd%3AQ2352616%20%7D%20.%0A%09%09%7D%20%0A%09%7D%20.%0A%09SERVICE%20wikibase%3Alabel%20%7B%20bd%3AserviceParam%20wikibase%3Alanguage%20%22en%22%20%7D%20.%0A%7D>
shows that the majority of instances of P972 as a property that _don't_
violate any database constraints also fail 'a' . The bulk belonging to this
list:
https://www.wikidata.org/wiki/Q5460604
Looking at how others had used the property on Wikidata did not help me
with a correct interpretation.
As you suggest, I'll follow up this up on the P972 talk page on wiki.
I thought there was a way forward here by correcting the usage of P972 to
enable project task lists by correctly representing them as catalogs. It
seems P972 has its own issues, and only works well for a narrow definition
of catalog. James' new property proposal is a better way, and likely
represents the intended function.
There are also issues of notability concerning many of the items
created. these should be addressed by evidence (i.e.
citations), but
instead we are being given argument by authority.
Really I thought the argument was more: "We know / have good reason to
believe these items are notable, please allow us time to demonstrate it (or
please join our effort and improve it yourself!)"
It seems clear
that the Black Lunch Table Wiki Project has a list of
artists
they are interested in, so I consider that the
Black Lunch Table catalog
is
real thing.
Does it meet our notability requirements? How?
Indeed does it exist? Where? In what form? How can I see it? If the
answer to these latter questions is "on Wikidata", then the argument
is circular.
I took 'Q2352616:catalog' to be an abstract entity as it is a subclass of
work, referring to the FRBR concept.
From FRBR definition of a work ( 3.2.1
https://www.ifla.org/files/assets/cataloguing/frbr/frbr_2008.pdf ):
"A work is an abstract entity; there is no single material object one can
point to as the
work." so my argument would be that where, how, or how many times it is
expressed does not affect its existence as an abstract Q2352616:catalog.
I do accept that those factors can be taken into consideration when
establishing notability.
Deciding how Wiki Projects or their task lists are notable probably comes
down to whether "3. Using Wikidata as a shopping list for a Wikiproject."
(point 3 from Maarten Dammers' email in this thread) is deemed an
acceptable thing on Wikidata.
BLT and these similar projects seem primarily "useful" independent of their
external notability. Whether "usefulness" automatically confers
notability, I don't know. It seems that these projects and this attempted
meta-project tagging is somewhat new, and we are all trying to clarify the
guidelines.
Perhaps this raises questions about whether stubs or
placeholder entries
are
acceptable on Wikidata?
We already have the answer; the notability policy. It is generally
very inclusive - far more so than Wikipedia's - so the egregious
failure of BLT community to demonstrate notability for many of the
items they are creating is telling.
This is one area that maybe needs clarification then, from the notability
guidelines page:,
2. It refers to an instance of a *clearly identifiable conceptual or
material entity*. The entity must be notable, in the sense that it *can be
described using serious and publicly available references*. If there is no
item about you yet, you are probably not notable.
That says the 'entity must be notable', and 'can be described using serious
and publicly available references' which I take to be quite different from:
'entity must be *shown to be* notable' and '*must* be described *(in its
wikidata entry)* using serious and publicly available references'. The main
criterion above for the purposes of entry content is the 'clearly
identifiable' one, and that can only be best efforts since any example of a
better populated entry could still be coincidentally ambiguous if someone
with the same name, from the same place had a similar occupation at the
same time. External sources are not infallible either, so even having
external ids does not guarantee there won't be potential confusion
(admittedly this would be a pretty pathological case, but consider
pseudepigrapha where there are differing valid opinions on identity of
authors).
I interpret notability criterion 2 as meaning that there will be entries in
Wikidata that are valid, but incomplete, and their notability is not
immediately obvious or demonstrated from the entry alone, yet that entry
could be useful to someone with external context who went searching for
it... and that this is completely acceptable in Wikidata.
If this is not correct, I think the Notability criteria 2. needs
clarification to emphasise the 'must be demonstrated' part. I think it's
already clear that demonstration is good and better, and that specific
entries can be challenged without it, but demonstration does not appear to
be a strict requirement for creating an entry according to that page. That
seems fine to me, I think insisting on perfect sourcing for every entry
before adding it would be impractical and make Wikidata much less useful.
A "poor" entry with a BLT, or any other project tag, is better than a
"poor" entry without one, because at least I know there is some project
that claims to care about that data, and I can follow up with them if I
can't figure out how to disambiguate the entry myself.
Aside: The last part "If there is no item about you yet, you are probably
not notable." seems unnecessarily snarky, and of questionable value in that
position, <shrug>.
Last thing,
apparently consensus advice was given to use P972:Catalog in
this way
So we're told. Where is it evidenced? I for one don't find any clarity
about what was intended, in the discussion in which I participated,
which was a proposal for a property for the supposed catalogue's
unique identifiers - identifiers which it now turns out do not exist.
As I indicated above, discussion is ongoing on Wikidata. I suggest
that further discussion take place there, rather than on this list, as
on Wikidata it will carry more wight in deciding consensus
Thanks Andy, it's clear there is confusion :) I appreciate that splitting
the discussion in more locations doesn't help eliminate that!
Sharing my understanding of what seemed like the clearest issue with the
catalog use was an attempt to help others by stating an argument clearly
and specifically.
Your response makes me agree that P972:catalog is inappropriate for this
use, and that a new property will be more useful. I'll add my relevant
comments on the recent proposal here:
https://www.wikidata.org/wiki/Wikidata:Property_proposal/Wikidata_focus_list
I think there are still problems with current P972:catalog usage/definition
(independent of the 'controversial' usage which started this thread) which
probably contributed to the overall confusion (certainly mine!) and I'll
add something on the P972:catalog page on wiki too.
Regards,
Charles.
