Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > sci.med.informatics > #18

Med Epistemology (Bayes, Burke & Barzun)

Show all headers | View raw

sci.med:556795 14 Aug 2016 

   I have come to realise that what keeps science from spurious results is a
priori encyclopedic knowledge.  Two Anecdotes. When my dad was diagnosed with
stomach cancer, I was told it was caused by nitrates in cold cuts. But when
the cancer came back, as the Viagra Nobel showed nitric oxide to be the
primordial mammalian neurotransmitter, I was told the 1930s theory of
Helicobacter Pylori had been revived.  In grand rounds a year ago, a young
resident presented a case of an 89 yr old with enlarged heart as genetic,
only to have an older doctor insist it was caused by 60 yrs of high blood
pressure. I asked someone at the genome center and he said adding Bayesian
priors to expert systems reduces spurious results. Bayes Law brought Kant's
concept of a priori to mathematical probability. I have always felt
uncomfortable with this concept of evidence based medicine precisely because
every few years a totally new theory throws out all previous knowledge and
makes new claims. Instead it pays to see why the old theory was wrong and to
learn from our mistakes, not totally dismiss them. This really extends from
the Hun or German peerless (hence uninspectable) hyperspecialist model of
education which now seems to overtake the anglohellenistic model of peer
reviewed encyclopedic (encyclios paideia) general education championed by
Barzun.  The peerless hyperspecialist might as well be a shaman or guru
shrouded in mystery.  I accept the errors of those who rejected Galileo,
although you could argue they weren't based on scientific study but blind
ideology, hence a different type of problem. I view ideology as the worship
of human hueristics and the cause of atrocity when the heuristics take
precedence to reality.  I believe Burke's central thesis, that change should
be measured and studied because if we replace everything at once, we will
have nothing to stand on, or pulling the wrong thread could unravel the
fabric. Further, as Sydney Hook warned, studying our old errors (or claptrap)
keeps us from repeating them. Before "Japanese Innovation" we were taught
that we should follow standards so we could continue to use old results,
algorithms and equipment.  As an example I learned electronic devices should
be designed to handle (fan out) five accessories, but the modern ones reduced
it the the bare minimum of one.  Everything is now designed to only work in
the short run, as if Keynes admonitions "In the long run, we are all dead"
has been extended to science and engineering.


				    - = -
 Vasos Panagiotopoulos, Columbia'81+, Reagan, Mozart, Pindus, BioStrategist
		    http://www.panix.com/~vjp2/vasos.htm
  ---{Nothing herein constitutes advice.  Everything fully disclaimed.}---
   [Homeland Security means private firearms not lazy obstructive guards]
 [Urb sprawl confounds terror] [Phooey on GUI: Windows for subprime Bimbos]

Back to sci.med.informatics | Previous | Next | Find similar | Unroll thread

Thread

Med Epistemology (Bayes, Burke & Barzun) vjp2.at@at.BioStrategist.dot.dot.com - 2017-02-10 22:54 +0000

csiph-web