Re: Letter to Shirley Jackson

[David Scherer <scherer@uiuc.edu>]

Dr. Cohen said:

> 	--As scientists, we are only involved with definition No. 1 [i.e., a
> correspondence definition of truth]; we contend that LNT has been proven 
> to be wrong, grossly over-estimating the risk of low level radiation. It
has 
> also been proven that the  basis for LNT, which is entirely theoretical, is 
> wrong. Discarding a theory which has been tested and has failed in that 
> test is at the heart of  "The Scientific Method". Not discarding it is a
violation 
> of The Scientific Method. 

1.  Scientists use a functional definition of truth (no. 3 in my previous
post) all the time.  We use classical electrodynamics even though we know
quantum electrodynamics is more accurate.  We use the ideal gas law even
though it fails in precise measurements.  In "The Scientific Method" ALL
theories and models are approximations, accepted tentatively to make
predictions and/or decisions.  These approximate models are accepted in the
domain for which they are developed.

2.  As I understand it, LNT is used as for safety purposes, to limit
exposures below hazardous levels.  Do most radiobiologists actually believe
that _all_ late, stochastic effects are precisely linear?

3.  Regardless which definition of truth one uses,   Dr. Cohen is correct;
all theories and models must be refined as our knowledge is improved.
Perhaps it is time to refine our model of radiation biological effects.
But a specific, numerical alternative must be proposed and tested.  It is
not enough to say there is a threshold, there must be consensus on the
numerical value of that threshold.  This requires positive evidence, i.e.
measured phenomena actually showing the emergence of biological effects at
a given dose absent at a lower dose.  (It is not enough to say that effects
have not been observed when the statistical power is inadequate.)  This
effect must be shown in radiation of different qualities (both low and high
LET), different exposure rates (acute and chronic), and using varied
investigational techniques (ecological, case-control epidemiology,
experimental biology, etc.).  I look forward to the day when we will have
this information.

Regards,
Dave Scherer
scherer@uiuc.edu