[Haskell-cafe] Do you trust Wikipedia?

Henning Thielemann lemming at henning-thielemann.de
Fri Oct 19 10:30:47 EDT 2007

On Fri, 19 Oct 2007, Jules Bean wrote:

> jerzy.karczmarczuk at info.unicaen.fr wrote:
> > *PLEASE*, show me untrustworthy Wikipedia pages.
> Any article on a disputed territory or open political dispute.
> Most articles on a controversial philosophy.
> Many articles on living people.

Articles on controversal topics like HIV/AIDS, climate change, economics,
generally things which are called "conspiracy theory" by enough people.
You find many authors which forget good scientific style if the topic is
presented in a way which contradicts to their view. "Neutral point of
view" just means "Biased view which is shared by enough people".

> Fortunately, articles on mathematics have several virtues which make
> them less likely to subject to the problems which plague the above.
> There is a notion of objective proof,

Most proofs in mathematics use intuitive arguments, most proofs are not
formalized enough in order to be checked by machines. Ok, this can be
considered a deficiency of machine provers. But in the history there were
famous "proofs" which turned out to be wrong. Remember the first "proof"
of the four color theorem of Alfred Kempe (cited from, you guess it,
wikipedia :-)  http://en.wikipedia.org/wiki/Four_color_theorem ) Or
remember the first trial of Andrew Wile to prove the Taniyama-Shimura-Weil
conjecture for Fermat's last theorem. It is generally hard to show that a
proof is incorrect if the statement is correct.

> there is a wide peer-reviewed literature which can be used as
> references,

Do referees actually check every proof?

> there are a disproportionate number of wikipedians with mathematical
> ability to catch errors.

Wikipedia contains the same abuse of notation as all the mathematical
literature. I just like to recall "the function f(x)". Or see the German
part of Wikipedia where people have decided to restrict the category
"Mathematical Function" to functions with real and complex parameters and
values. (http://de.wikipedia.org/wiki/Kategorie:Mathematische_Funktion)

In conclusion I would not trust Wikipedia. But this holds for the rest of
the world, too. Scientists must always have a basic stock of scepticism,
especially for "well known" and "widely accepted" facts/believes.

More information about the Haskell-Cafe mailing list