[Wikipedia-l] Analyzing and Visualizing the Semantic Coverage of Wikipedia

[Lars Aronsson]

Delirium wrote:

> This actually doesn't surprise me too much.  Hindi is a lingua 
> franca of sorts in India, but so is English.  In particular, the 
> Hindi speakers most likely to be editing an encyclopedia on the 
> internet (middle to upper class, well-educated) are highly 
> likely to also speak very good English.  Anecdotally, there are 
> indeed a lot of Indians making good contributions on the 
> English-language Wikipedia,

This is the case in Scandinavia too.  I would say that a majority 
of the Scandinavian contributors speak excellent English, even if 
the user base is starting to expand to older, less educated 
contributors.  Some Scandinavian wikipedians think the 
Scandinavian language versions are jokes, and only contribute to 
the English Wikipedia. But the more Scandinavians who contribute 
to the English Wikipedia, the more contributors spill over to the 
"native" languages, of which 4 are among the top 20 languages at 
Wikipedia. Danish with 5 million speakers has 47,000 articles.

So even if these supposedly young, male, upper-class wikipedians 
in India speak good English, that is not enough to explain why 
Hindi is lagging behind. Perhaps the issue is if they grew up with 
an English or Hindi encyclopedia in their parents' bookshelf.  Do 
they consider Hindi to be "a language of encyclopedias"?  I don't 
see the word "Hindi" in http://en.wikipedia.org/wiki/Encyclopedia
or http://en.wikipedia.org/wiki/List_of_encyclopedias

I also don't see the word "encyclopedia" in any of
http://en.wikipedia.org/wiki/Hindi
http://en.wikipedia.org/wiki/Hindi_literature
or http://en.wikipedia.org/wiki/Indian_literature

In the longer run, I hope Wikipedia should not only have a full 
good Hindi version, but one edited by the right proportions of 
gender, class, and age.  We probably have a very long way to go.


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