From: turpin@cs.utexas.edu (Russell Turpin)
Subject: A difference that makes no difference is no difference
Summary: When a parapsycological hypotheses becomes religious
Date: 24 Jan 91 00:47:24 GMT

-----
Descartes, in wondering what we can know, created a famous
gedanken.  What if, he asked, the world as we know it does not
exist, but instead, we are souls completely controlled by a demon
who fabricates all our sensory experience?  The modern version of
this gedanken places the subject's brain in a vat connected to a
gigantic computer run by an evil scientist, but the effect is the
same.  In both gedanken's, an ordinary view of the universe is
replaced by a more complex one, where perceived reality is based
on a deeper one that is perfectly hidden.  The key assumption in
both cases is the same: the demon or the evil scientist has
contrived (so far) to create experience for us that exactly
duplicates what it would have been if the universe were exactly
as ordinary perception leads us to believe.

Such a hypothesis is "bulletproof".  There can be no evidence for
it, but it is also immune to disproof.  There may be aspects of
reality that are totally hidden to us.  Most people, on
understanding that this is not such a profound claim, indeed,
that the opposite would be quite surprising, accept a pragmatic
resolution.  There is little point in speculation on the kinds of
things that if they are, are also perfectly hidden.  In
investigating our universe, we are practically restricted to
looking at the kinds of things that make a real difference in
what we can perceive.  (Descartes preferred a religious solution.
In essence, he assumed that the demon is real, but is a good god
rather than evil devil, and so we should place our faith in it.)

The acceptance of any such hypothesis is inevitably religious or
metaphysical (in the bad sense), and in my opinion, needless.  It
certainly lies outside science, history, and every other kind of
empirical study.  Despite this, bulletproof hypotheses are often
put forth to defend a putatively empirical position.  Perhaps the
most famous case of this occurred with the first clash between
creationists and scientists in the 19th century.  Why, asked a
preacher whose name I forget, does the universe show evidence of
vast age, when it was created a mere few thousand years ago?
Because, he answered, the Christian god is perfect and would not
create anything without giving it a history!  The universe was
purposely created to give evidence of vast age, even though it is
young.  This "theory", named Omphalos because it answered that
famous question about Adam, was quite popular for a while.
Philosophers critized it, and pointed out that it was precisely
the same kind of hypothesis as the existence of Descarte's demon.
Indeed, such a god could have created the universe a mere five
seconds ago, replete with our memories of yesterday!  (Today,
this argument is viewed as mere sophistry, except, as one might
expect, among certain creationists.)

A more subtle way of bulletproofing beliefs is to let them make a
difference, but only in special and guarded fashions.  Thus, say
some believers, there really *are* miracles and people really do
see them, but only believers.  Miracles never appear in a fashion
that permits their common study, because god will *not* be
tested.  It is not hard to see that this is another bulletproof
hypothesis, and leaves belief in miracles a matter of faith or
special dispensation.

One can empirically investigate the age of the earth, but one can
also bulletproof one's hypotheses about time and history so that
they become solely a matter of faith.  The same is true with
parapsychology.

Hypothesized abilities such as telepathy and precognition could
exist in a way that is clearly testable, even if only few people
experience them, or they appear only sporadically, or only in
times of stress, or only with practice, or only when not
previously used, etc.  But it is also possible to bulletproof
hypotheses about such phenomena, in much the same way as it is
possible to bulletproof hypotheses about the age of the earth or
the existence of miracles.  "Precognition insures that telepathy
occurs only when it cannot be tested." "Parapsychic effects make
sure that only those who believe in them can see the evidence".
Etc.

When a hypothesis is bulletproofed in this fashion, it is moved
beyond the realm of empirical study and shared investigation.  We
can no more "expand the scientific paradigm" to study such
hypotheses than we can somehow empirically determine the number
of angels that strut on pinheads.  Those who suggest such things
show only shallow philosophy, their misunderstanding of science,
and the pernicious desire to disguise *their* faith as a matter
for empirical study.  Parapsychologists who bulletproof their
hypotheses, who I hope are few, are practicing sophistry, not
science.  They do not need a new "paradigm"; they need to
understand philosophically what it is that they have done.

Russell