Logical fallacy

["ruth_weiner" <ruth_weiner@email.msn.com>]

Dear RADSAFERs:

Please, if you have time, comment on the following thesis for me.  I want to
find out if I am totally off base.  (This is probably more appropriate for
RISKANAL, but I am not on RISKANAL any longer.)  Here is the thesis:

There are two logical fallacies associated with the linear non-threshold
theory (LNT): one in the theory itself and one in its application to the
carcinogenesis of radioactivity.  (I am trying to figure out what these
fallacies are called in logic nomenclature).  The fallacy with the LNT
itself is the
fallacy involved in any extrapolation back to zero.  The fallacious
statement is: because a threshold has not been positively and definitively
demonstrated, we may logically assume that there isn't one.  In fact, one
cannot deduce the absence of a threshold from the failure to observe one.
An example is CO.  The essentially linear dependence of effect on dose was
well-known long before the threshold had been observed.  However, there is a
well-established threshold for CO toxicity (it is the fraction of of
carboxyhemoglobin that is produced by cell respiration), and neither its
existence or its value could be deduced from the linear dependence of effect
on larger doses.  In fact, this type of deduction is not peculiar to
linearity.

The second type of fallacy is in the application of the LNT.  There are, in
fact several fallacies in application.  One is that the deduction leads to
an absurd result.    For example, if one applies the EPA notion of the
toxicity of inhaled diesel exhaust, weighted average 3.08E-8 fatalities/km
(Biwer and Butler, Risk Analysis 1999, p.1166), to the approximately 200
billion truck-km traveled annually in the US (DOT web page) we could expect
6000 deaths each year whose ONLY cause is inhaling diesel exhaust from a
passing truck.  Carrying this further, a cyclist who cycles about 10 km a
day on city streets has a lifetime (70 year) risk of 7% of dying from
inhaling truck exhaust -- about the same order of magnitude as his or her
risk of dying in a traffic accident.

The second fallacy in application is illustrated by the old chestnut about
the inability of nine women to produce a baby in one month.  More to the
point: if one person inhales a liter of water in one lungful, he (or she)
will drown.  However, if a million people each inhale one microliter of
water in one lungful, one of those million will not drown.  If one person
swallows 50 aspirin tablets, that person will suffer kidney failure, but if
50 people each take one aspirin, none of them will suffer kidney failure.
This translates to the person-rem = person-rem fallacy.  One person
receiving a dose of 50 rem has a different result from 10 people each
receiving an average dose of 5 rem, and that is different still from 1000
people each receiving a dose of 50 mrem.  Of course, if one of those 10 or
1000 receives most of the 50 rem, both the dose and the result would be
about the same as one person receiving 50 rem.

Please note that I have not addressed whether or not there is a threshold
for radiation damage -- that is entirely aside from the concerns expressed.

If you don't want to clutter up RADSAFE, or if Melissa thinks this is too
far off-topic, please reply to me privately.


Ruth F. Weiner, Ph. D.
7336 Lew Wallace NE
Albuquerque, NM
505-856-5011
fax 505-856-5564
ruth_weiner@msn.com









************************************************************************
The RADSAFE Frequently Asked Questions list, archives and subscription
information can be accessed at http://www.ehs.uiuc.edu/~rad/radsafe.html