[PREV - RUSSELL_ALL_TRUED_UP]    [TOP]

RUSSELLING_THE_DOUBT

                                             May 25, 2022


    Bertrand Russell again
    writing about Dewey's              RUSSELLING_DEWEY_1939
    "Logic: The Theory of
    Inquiry", p.174 of his                      BASIC_RUSSELL
    "Basic Writings":


    "... Nor would he reckon suicide a suitable method, although it
    would be eminently effective in removing doubt. We must therefore
    ask ourselves what he can mean by ‘need for doubt’."

    "For those who make ‘truth’ fundamental, the difficulty in question
    does not arise. There is need for doubt so long as there is an
    appreciable likelihood of a mistake. If you add up your accounts
    twice over, and get different results, there is ‘need for doubt’;
    but that is because you are persuaded that there is an objectively
    right result. If there is not, if all that is concerned is the
    psychological fact of inquiry as an activity stimulated by doubt,
    we cannot lay down rules as to what *ought* to remove the need for
    doubt ..."


Let's run with Russell's
example here a bit.  I have
some experience with crunching    As an aside: it's always preferable to break
through long calculations and     down long calculations into shorter ones, and
trying to gain some confidence    to determine intermediate results that can be
that you've done it correctly.    checked independently.  You can't *always* do
                                  this easily--

                                        I'm thinking of the days when I needed
                                        to apply a formula from "Formulas for
                                        Stress and Strain" without access to a
                                        computer you could use to just turn
                                        them it a program.

What happens in practice is you may
find you need to do the same
calculation a half dozen times or
more-- you may get two or three
different versions of the result and   That process *does* presume that
you need to keep re-doing it until     there's only one correct answer, and
you're certain which one of them is    that particular belief comes out of
in the majority.                       long experience with arithmetic
                                       calculations-- and perhaps a
The central point: even in this        theoretical understanding of the
remarkably simple case, there's        logic underlying the mathematics.
a probabilistic character to the
proceedure, there's room for                   And in this case, this doesn't
some doubt about whether our                   rely on a belief in an
criteria for eliminating "need                 objective, external reality,
for doubt" is adequate.                        but rather the internal
                                               consistency of our mathematics.
There could, for example, be a mistake
that you're making repeatedly, perhaps
because some quirk of the process is
creating a psychological trap you keep
falling into-- you're putting decimal
places where you feel like they should
go and not reading closely enough,
you're punching a closing parenthesis
in at the wrong stage and grouping
intermediate calculations wrong, etc.

The majority result could still be wrong,
and even calling in a second person to
re-do your work might not be an adequate
check, they may hit the same psychological
trap that you did...

Like it or not, there's a pragmatic tradeoff
here between degree of certainty and the
amount of work you're willing to do.


   Sitting in the corner and chanting
   "I believe in *truth*, I do, I do."         RUSSELL_ALL_TRUED_UP
   doesn't actually alleviate all doubts.


      I think that Russell may have fallen into the trap
      of thinking that there's something inherently
      superior about thinking in the ways he's used to
      thinking-- he's confusing the actual "need for
      doubt" in the intellectual process under discussion
      with his own feelings of doubt while he's trying
      to think about it.



--------
[NEXT - RUSSELLING_DEWEY_1939]