X-Message-Number: 8245
Date:  Mon, 26 May 97 13:22:00 
From: Mike Perry <>
Subject: Pancritical rationalism, testability, falsifiability

Tim Freeman, #8241, has some interesting arguments as to how, in 
practice, we might avoid having to adopt unfalsifiable hypotheses. 
His method is based on pancritical rationalism, in which choices of 
actions and plans are evaluated, in addition to beliefs. We can 
focus, then, on our behavior and not have to take a stand on whether we 
believe something that is not provable or disprovable. He offers an
argument for "valuing tomorrow's consequences" of one's actions today,
which is a de facto rejection of the day-person hypothesis, without
explicitly referring to it. This argument is based on evolutionary 
considerations: the species will thrive better if we in fact (and 
humans in general) choose this course rather than acting as if our
lives will end as soon as we sleep.

I think this is a worthy effort, but I have some criticism.
Pancritical rationalism may provide a way to 
avoid unfalsifiable hypotheses, but I think, in this case, it has
resulted in an unnatural approach. (It might be applied differently 
to come up with a different rationale, however.) To simply take
the stand: "I *do* survive a period of unconsciousness, I *will*
be present in the world tomorrow (barring something more drastic
than sleep), therefore *I* have a direct, vested interest in the
consequences of my actions today" 
seems much more reasonable to me. (One thought that goes along with 
this is that I am not really much interested in "survival of the 
species" unless I will participate in that survival, directly.) But this,
in turn, invokes the "continuation" hypothesis (my 
terminology) that the day person-hypothesis is false, i.e. that 
we do, in fact, survive periods of unconsciousness.

Neither hypothesis is falsifiable, we might say. However, in 
actuality, from *my* perspective they are falsifiable, based on my 
understanding of what a person is. That depends in turn, however, on 
the information paradigm, on regarding a person as like a program 
running on a hardware device, that can be restarted after a period 
of inactivity or "shutdown," etc. On that basis clearly the 
continuation hypothesis is valid, and the day-person hypothesis is 
not. (Presumably, then, I would not be inconsistent with pancritical 
rationalism if I rejected the day-person hypothesis outright, given this 
view.) But not everybody will see it that way, or some other way that 
would resolve the question, so for the world at 
large it seems that both hypotheses must be considered unfalsifiable.

There are other hypotheses that, while unfalsifiable in a strict 
sense, seem at least to be partially testable. The claim that 
supernatural powers or forces exist is one such hypothesis. Though 
perhaps it can never be fully tested, I think the scientific evidence 
against it is good, and I feel justified in rejecting it. I take my 
rejection as a "working hypothesis" and not a dogmatic certainty, and 
remain open to any evidence that might decide the question either way.
More generally, I try to replace "beliefs" by "working 
hypotheses" to remain open to whatever the evidence will show, while 
at the same time enjoying the advantages of taking a definite stand.

Then there is the sort of hypothesis, such as that Perry Metzger's 
invisible bunny exists, which might be said to have really nothing
going for or against it, in the way of experimental evidence. In
other words it is truly, totally untestable. But this can be thrown
out on grounds of Occam's razor, if nothing else. 

Mike Perry 

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=8245