[Wikipedia-l] Are we running out of sources

[Lars Aronsson]

Thomas Dalton wrote:

> Primary sources are useful, but only for certain things. 
> Linkedin could never be the main source for an article.

I think everybody agrees. But it is still interesting to ask: How 
would we cite information found in Linkedin?  Via some web archive 
service?  And have there been any widely reported cases of 
Linkedin fraud, where somebody listed a PhD title that they really 
didn't have?  Or is that something very common?

> What makes you think people didn't lie?

I can only speak for the Swedish Who's who ("Vem är det"), which 
has been published 45 times between 1912 and 1999 (roughly every 
second year), with the 46th edition in 2007 (an 8 year gap and 
change of publisher, it was discontinued and then revived).  I 
haven't heard of any cases of fraud.  Criticism was launched 
against the 2007 edition, that it lacked several business leaders 
and the prime minister, since people were allowed to opt-out.

The Norwegian ("Hvem er Hvem") has been published only 15 times 
between 1912 and 2008. The most recent edition only listed 1000 
people, but editions 1-14 listed between 3200 and 4900 people.

The Danish ("Kraks Blå Bog") has been published every year since 
1910, with recent editions listing some 8000 people, meaning that 
one in every 700 citizens (5.5 million Danes) is listed.  Among 
these three countries, the Danes have the best reference works 
(closest to Germany), but the smallest version of Wikipedia.

It was very interesting to hear about the Russian version, and its 
problems with post-Soviet denial.


What I'm coming to is that Wikipedia might have to adopt the 
method of sending out forms to select people, asking for their 
biographic details (or for verification or denial of what's 
already in the Wikipedia article). That doesn't mean we should 
trust such autobiographic information blindly, but allow this 
input in a controlled form to make Wikipedia more complete without 
encouraging the uncontrolled editing of your own article. Such a 
suggestion of course begs many questions, e.g.:

 * Who should send out the forms?  How do we introduce ourselves?  
   How do we explain that all of Wikipedia's rules still apply 
   (no, you can't opt out; no, I can't edit to your favour), to 
   people who never thought of editing Wikipedia, and might have 
   no idea how it works?

 * How should the received forms be stored and referenced?

 * If we discover false claims or grave omissions in the received 
   forms, how do we handle the next contact with that person?

 * Should the input perhaps be handled as interviews for Wikinews?  
   Somebody can do a detailed interview for Wikinews, and then 
   Wikipedia can cite (parts of) that interview.  Does that scale?  
   We would have to explain to each person what Wikinews is, but 
   perhaps "an interview" is easier to understand than "a form 
   from Wikipedia" (the latter sounds like "you have won a lottery 
   from Microsoft", the typical spam scam).

In the case of Who's who, it's of course the editors employed by 
the publisher that sends out forms asking for details. This only 
highlights how completely different Wikipedia is.

It would be interesting to hear if anybody tried something like 
this already, perhaps within a limited Wikiproject?


-- 
  Lars Aronsson (lars at aronsson.se)
  Aronsson Datateknik - http://aronsson.se