Dear Marion,

Thank you for your letter. I'm much appreciative of the seriousness
of your reflections about the statistical medical evidence on which
contemporary physicians base their therapeutic decisions. Having 
spent much of my professional life advising my patients how much
medication, if any, they should take to "control" numerically
calibrated intraocular pressures, relying in part on a mathematical
model of my own devising:
http://home.earthlink.net/~ernstmeyer/glaucoma/letter02.html
http://home.earthlink.net/~ernstmeyer/glaucoma/letter03.html
- I've probably sent you this URL before pointing as it does
to a mixture of "serious" science, parody, irony and satire -
I'm in no position to be critical and have much to apologize for.

There appears to me in statistical modeling of experience an 
"uncertainty principle" conceptually distinct, yet quite analogous
to the uncertainty principle in the quantum theory of physics.
The primary statistical uncertainty is in the translation of
experience (Erleben) into logical or mathematical symbols (words
or numbers), a process in which sometimes little, sometimes much,
occasionally everything is lost.  A further statistical uncertainty
attends the grouping of individual phenomena, - events, persons, -
into classes, again a process in which sometimes little, sometimes
much, occasionally everything is obscured. The certainty exhibited
by statistical expressions is never a reflection of primary
experience. Statistical conclusions come to represent a self-contained,
self-validating reality of their own, a latter-day logical-mathematical
scholasticism which is likely to control human activities for
centuries to come, not only in medicine, but in the basic and not
so basic sciences, such as economics, as well.

Suddenly, I find myself sleepy, having gotten up at 6 a.m. to drive
Klemens to his office. My thought is becoming unthreaded.
I need to catch up on my sleep. Maybe more later.
Jochen