| Re: Ockham's Razor against unfalsifiable hypotheses |
|
 |
|
 |
|
 |
|
 |
Group: alt.atheism · Group Profile
Author: David SchwartzDavid Schwartz Date: Mar 9, 2007 05:44
On Mar 9, 2:06 am, mike bishop yahoo.co.uk> wrote:
> I understand the gist of what's being said here, but I always thought that
> Ockham's Razor couldn't be used against unfalsifiable hypotheses, in which
> case we *would* have to withold judgement on the fairy thing, even though
> that's undesirable. Am I wrong?
Ockham's Razor is equally applicable both to unfalsifiable theories
and difficult to falsify theories. You can blue the boundary if you
like, consider the following thought experiment:
You want to demonstrate experimentally that gravity is a 1/r^2 force,
so you make many, many measurements. When you finish, you find that
all your measurements support the 1/r^2 hypothesis. However, there
must be some maximum distance at which you tested, regardless of how
many tests you have done.
So, what happens when someone hypothesizes that up to that distance,
it's 1/r^2 but after that, it's 1/r^3. Doesn't your data equally well
support both hypotheses?
Technically, this particular hypothesis is falsifiable, just test with
a larger distance. But a new hypothesis, 1/r^3 after *that* distance
is right on the horizon.
> Also, assuming that I *am* wrong and that the razor takes out the fairies,
> would it be possible to make an inductive argument against the fairies: an
> argument concluding that probably there are no fairies in the garden?
So long as the fairies have no testable consequences, I don't see how.
DS
| 
