On Thursday 23 February 2012 01:10 AM, Thomas Morton wrote:
Splitting this off, Achal, I hope that's OK :)
There's a discussion on at the reliable sources notice board, for
instance,
Thanks for the pointer there; I'll try and place some comments in there
later. It is certainly an interesting discussion.
But here are some initial thoughts (please bear in mind I have only
scanned
that discussion, and whilst I have had an ongoing interest in the oral
citations project I never dug into in too much depth). Also remember this
is based on my interpretation of our policies, so others may well differ!
Can I ask you how you would analyse the work of
the oral citations
project
(
http://meta.wikimedia.org/**wiki/Research:Oral_Citations<http://meta.wik…)
in terms of our policies on original research, and verifiability?
The best way I can address this is to lay out my thoughts on our sourcing
policy.
Material on Wikipedia can be divided into "fact" and "opinion". The
latter
of these is, perhaps confusingly, the simplest to address; because
opinion,
viewpoints and perception can quite easily be collated and summarised.
The
only real difficulty exists in figuring out which opinions are noteworthy
to record.
The problem is facts; as I am sure everyone can appreciate, facts are
very
easy to get wrong (maliciously or not). This is especially a problem in
History where events can be pieced together via all manner of sources.
Even
WW2 history can differ dramatically depending on the accounts you read -
some overuse oral citation (humans are fallible) and others misuse
official
records (which can range from faked through to inaccurate).
The problem with primary sourcing of the oral form is that it comes
directly from an individual - with all of their perceptions and
biases. To
make an extreme example out of this; imagine taking an oral citation from
Hitler, and a Jew in a concentration camp. Such citations would, I
imagine,
give radically different viewpoints of the Holocaust. Obviously other
accounts, by third parties, show us which account is accurate - but if we
had only those two viewpoints I hope it is obvious how
difficult separating fact and fiction could be (ignoring that any
rational
person would see the obvious).
Of course. So, for the oral citations project, we specifically chose
topics that are in the present, that are seen and done by thousands of
people (i.e. not obscure), and that are also as uncontroversial as
possible. Examples: village games, temple rituals, recipes.
So that brings us to the ideas behind sourcing; which is that we should
consider not only the material but author and publisher. This is
important
because if the author of the source is partisan to the material then you
have to consider they may be biased to their viewpoint. As less extreme
example might be two citations from a Republican and a Democrat. Both say
"My Party is the Best because our policies are..." - you can't use either
source to say one party is better, because they are partisan. But you
could
use it to relate their parties policies; and as partisans they are well
positioned to relate those policies!
If the author is a third party, of course, that lends weight to their
material.
The publisher is the stumbling block in this case; because it is a
non-expert [sic] researcher uploading material to Commons. What could
mitigate this is a detailed description of the methodology used to
collect
the citations, which would allow editors to review it for problems.
One final thing to consider is that WP:V talks about controversial or
challenged material. Whilst that might be a risk policy on the face (it
would be easy to present something non-controversial but also not true as
fact) it's critical to letting us actually write article (otherwise we
would be stifled in citations :)). For example; I've sourced material to
personal sites before with minimal problems - sometimes it is questioned
and what I usually say is "If you can show someone saying the
opposite, or
make a sensible argument against, then lets remove it". (FWIW, and
this is
an aside, I think is relaxed form to building articles is a Good
Thing, and
we should do it more often - worrying about being wrong is stifling).
So now I've picked it apart here is my thinking; Oral citations on
Commons
could be excellent sources in the right context.
:)
Sure if the material is disputed or otherwise problematic then it is
better
to look for a source that has peer review. But for simplistic, factual
things then I think it is rock solid. One example that comes to mind
(and I
don't know if the Oral citations covers this sort of thing) is this: I
was
recently on holiday in New Zealand. They have excellent museums there,
fanatically maintained (which is amazing compared to most countries...).
One nice feature is that a lot of the Mauri history exhibits have audio
from those of Mauri ancestory describing some cultural fact or other.
Things like recipes.
Yes, we encountered exactly this. When looking for aspects of everyday
life that people both widely knew about and did in India and South
Africa, but were also undocumented in scholarship or even print,
everything we had came back to 'culture'.
Something like that is an excellent oral
citation; the author is
authoritative (being Mauri) and the content uncontroversial.
Boiled down, I think that oral citations have a distinct place as a
source
- and we should encourage people to consider them as sources when
writing.
But they are not something you could, for example, base an entire article
around. We should also explore ways to make them more "reliable", and
more
usable. For example making them obviously available to experts in the
field
they cover.
I agree that an article wholly based on oral citations is less desirable
than an article that mixes traditional and non-traditional sources. Two
things here though: (a) There are actually a pretty large number of
things that are both widely known and done and not documented in print,
so this is a real problem; and in these cases, having articles solely
based on oral citations could still be useful, akin to a stub, awaiting
further refinement. (b) Though this is not directly related to the
conversation, it does relate to earlier points made by Sarah and you
regarding primary sources. It is sometimes hard to separate fact from
opinion within the oral citation - which is to say, X person's fact may
be Y person's opinion, etc. - and in my opinion, even the cleanest set
of facts gleamed from an oral citation will contain some perspective or
opinion. I don't see that as a problem (and this is regardless of how it
would parse through the OR policy) as long as the perspective is
attributed as just that, or even challenged. We encountered this, and
recorded it - in articles on village games in Limpopo in South Africa.
The older ladies we spoke to said young people didn't play the games
they had just shown us, and the young people we spoke to said they did,
but with a slightly different template, and we recorded and reported it
exactly as said, as two conflicting perspectives. (Would that be a
responsible use of primary sources? I think so.)
It *is* important to get secondary coverage of a topic, because we are
tertiary source. This is the core idea of our primary source policy;
if we
utilise primary material and research something to the extent that we are
the main authoritative source that becomes *hugely* problematic!
And further, how these policies might apply to the idea of social
media, as
well as more private archives, say, corporate archives, being used as
citations? (And on that point, is there a difference between the the
Native
American folk archive at the Smithsonian and the corporate archives
of the
Michelin corporation in France, for our purposes?)
Corporate archives I would deal with in the same way as any primary
source;
use it to cite facts, bear in mind the author/publisher. WP:SPS talks
about
being wary of unduly self-serving material, and I think that is an
excellent way of putting the approach to corporate archives.
I had an interesting discussion on this with Florence and Christophe, so
I'll share this with them if they miss it here. My own first instinct is
to trust a self-avowedly 'neutral' source (like the Smithsonian) more
than a corporation (like Michelin) but for our purposes, it doesn't seem
to make much sense to treat them any differently.
Bear with me, I'd love it if the Michelin Corp was opening its archives :)
But having worked for them and living in the city of their headquarters,
their discretion and love of confidentiality is such that I doubt it will
happen :(
The situation is complex. Should corp archives be somehow trusted or not
much or not at all. I would say "it depends".
It depends on the company (reputation).
It depends largely of which department produced the archives. Docs produced
by marketing departments should be taken with a HUGE pinch of salt. The
language is non neutral, they conveniently drop the embarassing facts, and
they tend to forget to put basic stuff such as dates ("ok, it is written you
produce 10 millions yoghurt, but when was that ?") or references to
countries ("ok, you write that you sell 10 millions of yoghurt in Global
South, but can you better define which Global South you are talking about
?") (yeah, true stories even if figures are invented).
Docs produced by departments of research or finances, I would put a lot of
trust in them. There is always the bad luck to stumble on a cheating company
just as it also happens that Museum Staff host a black sheep from time to
time. But generally, I consider information out of these departments quite
safe.
But the most difficult ennoying point is simply that most corp archives
appear to be a mess. Because companies are bought and sold, information is
lost on the way. Because of poor communication between departments. Because
staff come and go. And because the acceleration of business processes
unfortunately make it so that in the past dozen of years, less and less time
and money has been spent (invested) on a proper archive system, on good
procedures and efficient implementation. So when you ask "can you retrieve
the past 20 years of sales regarding this yoghurt", you'll get a blank
stare. Truth is, no one knows the date and no one knows where to find the
info.
Some companies sometimes hire external services (private historians) to
"clean up" their archives and some good stuff can get out of this, such as a
book or a museum (Michelin did that. Do visit the museum
if you happen to come. It is very nicely
done).
Usually, I recommand good sense. If the information does not appear "weird"
or "controversial" at all, I use the corp information as "trusted
source".
If it is clearly misleading or potentially illegal info, I trash it. But in
between there is room to accept the data as long as there is another source,
that may not be so great but that appears independant. For large companies,
there are usually independant sources. But for most medium size companies,
not. I give the situation a certain degree of tolerance.
Difficult to put that into any sort of policy except for "good sense".
Florence
In terms of social media, this is tricky. Because social media is vastly
more accessible than other mediums - particularly to hacks. Wordpress
blogs
are trivial to make, for example, and you can sound authoritative or
convincing on a subject to a layman with only medium effort. I'd treat
these with more caution.
Phew, that was dumped out in a stream of conciousness way - so it
might be
a bit "buggy". But that's what I figure :)
Tom
(Just as a note; I consider "publisher" quite broadly - i.e. the
person who
hosts or maintains the material)
_______________________________________________
foundation-l mailing list
foundation-l(a)lists.wikimedia.org
Unsubscribe:
https://lists.wikimedia.org/mailman/listinfo/foundation-l
_______________________________________________
foundation-l mailing list
foundation-l(a)lists.wikimedia.org
Unsubscribe:
https://lists.wikimedia.org/mailman/listinfo/foundation-l
_______________________________________________
foundation-l mailing list
foundation-l(a)lists.wikimedia.org
Unsubscribe: