[WikiEN-l] Archives as sources proposal

[Fastfission]

On 4/14/06, Oskar Sigvardsson <oskarsigvardsson at gmail.com> wrote:
>
> Honestly, the verifiability argument doesn't fly at all with me. If
> you pick information out of an archive, it's most certainly
> verifiable, someone else can check it out as well. I realise that that
> is a hassle, but that doesn't change the fact that it is verfiable.


"Verfiable" of course does not mean "absolutely verfiable" (in which all
facts would at some level be), but clearly something on a spectrum from
"anyone can see it by clicking a link" to "it's an old book but there are
copies of it in at least a few major repositories where a large number of
Wikipedians are located." In the last category, it is worth noting that a
significant number of Wikipedians are students at universities with
extensive library services, including interlibrary loan, and it would not be
very hard to verify 90% of all very-old book references. It would be a rare
source indeed that somebody in the University of California system or in the
Ivy Leagues would not be able to verify within a few days, to speak nothing
of all of the other universities in the world with fine and extensive
library services.

If something is archival and unpublished though, it generally means that it
is in one single archive. That archive may have no interarchival lending
options. There may be no interested Wikipedians with access to that archive.
There are no doubt a few exceptions, but so rare would they be, and
certainly not worth changing a very good policy, one which sits as a chief
cornerstone to our epistemological approach, in order to satisfy. All of the
proposed benefits (a few birth/death dates? information about obscure and
probably non-notable institutions?) seem a paltry thing in comparison with
the fact that removing "verfiabiilty" would create an endless mess of
difficulty with POV-pushers.

How about, as a compromise, if someone would like to include a source
> from an archive, they are required to submit a copy of the document,
> so everyone can see it for themselves.


If they can host it semi-permanently somewhere (a scan at Commons, for
example), then it is, so far as I can reason, "published" and verfiability
is no longer an issue (NOR might still be, but that's a different can of
worms). If they cannot put it somewhere where editors of the future can see
it then no, it is still unpublished and unverifiable except to the small
group of editors who happened to see it at one point in time.

The fact is, it's really hard to define research in this context. How
> is looking up someones birth certificate in an archive worse than
> looking it up on the internet?Technically, they're both research.
> Honestly, I'd prefer it if the info got directly from the source. Same
> thing with things such as trial transcripts, if we quote someone from
> a trial, I'd prefer it if we knew excactly, word by word, what they
> were saying instead of trusting a third party.


NOR would not prohibit people quoting from trial transcripts or other
primary source material, so long as it was not being used in a way which is
"unique." I could quote all I wanted from the Oppenheimer trial but if I
quoted out of context in order to prove he was a space alien, that would be
NOR. If I did it to substantiate and add color to the standard
interpretation (or any of the possible notable POVs included in the
article), that would be just fine, of course.

The above only holds true if the primary source is accessible and verifiable
though. If I say that a quote in a secondary source is wrong but can't
provide any proof of it, then other editors are going to have to take me at
my word. That's a bad policy -- there are far too many people who either
purposefully or unintentionally mis-read or mis-quote to support their
arguments to allow this. Even I sometimes make transcription mistakes, and I
do archival research for a living. Perhaps it is not assuming "good faith",
but I don't want a policy which lets people assert things that I cannot
check. The entire system of Wikipedia relies upon multiple eyes being able
to see something, and what you're suggesting is that we just jetison that in
the case of primary sources (which are already difficulty enough for people
to interpret).


As long as you don't do analysis, and pick your information straight
> from the source, document it and be sure other people can see it too,
> what's the big deal? How is that any different from looking up
> information on the internet or in a magazine from half a century ago?
> Especially so if we make them provide a copy of the document.


If other people "can see it too" and we can "provide a (long term) copy of
the document" then verifiability is not a problem at all. There is no such
thing as using a primary source without "doing analysis", as any historian
will tell you. All reading involves interpretation. All quoting (and all
narrative) involves being selective. The current policies allow for plenty
of flexibility in regards to editing and researching; they require that at
the end, all interpretations are anchored in a secondary source, and that
all sources used should be verifiable.

I've yet to see a compelling reason why that shouldn't remain the case; it
has worked pretty well so far, and yet even now we have constant problems of
people inserting dubious and incorrect information into articles. At least
at the moment we have a strong argument for consistently removing
information which cannot be verified -- you would remove this check? What
benefit could possibly outweigh such a deficit?

FF