Re: Birkhoff Causation is Consistent With Rare Events But Not With Others

From: Osher Doctorow (mdoctorow_at_comcast.net)
Date: 10/07/04 

Date: Thu, 7 Oct 2004 11:57:04 +0000 (UTC)

On Wed, 6 Oct 2004 12:32:18 -0700, David A. Heisaer wrote:
>-------------------------------------------------------------------
>No, statistics is moving in a lot of directions. Please read the
current
>issues of the Journal of the American Statistical Association.
>DA Heiser
>>
>>
>>

This reply reminds me of a person who dislikes Impressionist paint-
ings criticizing an Impressionist's impression that art is going in
the Impressionist direction: "No, it's going in many directions.
See the current journal...."

By the way, my choice of Impressionism
as of Claude Monet is no coincidence. One of the greatest schools
of painting, it eventually lost popularity to "Modern Art," based
on concensus of "experts" (where are they now?) including journal-
appreciated "experts".

I anticipate the rejoinder that mathematics is not art, to which I
reply that as a mathematician I nevertheless recognize that classic-
al musicians like Beethoven and Mozart were among the greatest
Creative Geniuses of all time and so "art" and "humanities" can not
only be as great as any other field but are as great!

Of course Heiser has the germ of an idea in that there is enormous
dislike of Large Deviations and/or Fat-Tailed distributions from
many statisticians and probability people who prefer to ignore
the differences between/among Rare Events, Fairly Frequent Events,
and Very Frequent Events, and likewise for Events replaced by
Processes. In fact, if most probability-statistics people were to
recognize such differences, there would be a major bureaucratic
change in Academia and the Corporate and Governmental worlds
roughly similar to a nonviolent revolution - in my opinion. Notice
that Heiser did not say "in my opinion" in his criticism, which I
am often careful to do, though I don't always anticipate strange
replies since I am not a professional bureaucrat.

Osher Doctorow

Relevant Pages
Quantcast