Re: Small probabilities

From: George Murphy <>
Date: Mon Nov 07 2005 - 07:58:19 EST

----- Original Message -----
From: "Glenn Morton" <>
To: "'George Murphy'" <>; "'Bill Hamilton'"
<>; "'Alexanian, Moorad'" <>
Cc: <>
Sent: Sunday, November 06, 2005 11:16 PM
Subject: RE: Small probabilities

>> -----Original Message-----
>> From: [] On
>> Behalf Of George Murphy
>> Sent: Sunday, November 06, 2005 9:12 PM
>>But what a combination of quantum theory &
>> general relativity says is that lengths & time intervals smaller than
>> these
>> amounts can't be measured. That's not exactly the same thing, unless one
>> is
>> a convinced positivist.
> There is a parallel here with the multiverse. It too can't be observed and
> thus we can never actually say that all these infinitude of universes
> actually exist. If one can't measure a distance shorter than the planck
> length, on is then in the same situation. One can believe that they are
> there but one can't prove that they are there with any observational data.
> And that makes such a belief faith not science. And that is why I find it
> strange for the atheists who are strong proponents of the multiverse to
> claim that what they have is science and not faith.

I don't think the situations are quite the same. Classical general
relativity starts with a continuous space-time, & there's a sense in which
such a manifold can't be approximated by a discrete one. What you find when
you include both the QM uncertainty principle and the effect of energy on
clock rates in an attempt to measure a time interval with some (idealized)
clock is that when you get down to the Planck scale the net uncertainty in
your clock reading becomes comparable with the time interval that you're
trying to measure. In a sense what this means is that there is a continuum
of points but that the distances between them can't be measured with an
uncertainty less than 100% below the Planck scale. This is, admittedly, a
somewhat old-fashioned approach & what a lot of people working of quantum
gravity now are doing is start with discreteness (loop or string theories) &
then get classical GRT at large scales.

On the more general question - to paraphrase our ex-Pres, it depends on what
"observe" means. If the consequences of multiverse theories explain
observable phenomena better than other theories then they have to be taken
seriously, even if the other universes aren't directly observable. & one
also has to ask whether they're unobservable _in principle_ or just due to
our situation. E.g., in Linde's "bubble" version of inflation we could (I
think) observe another bubble if we were close enough to the edge of ours,
but the conditions that make life possible in our expanding bubble make that
extremely unlikely.

Received on Mon Nov 7 08:01:02 2005

This archive was generated by hypermail 2.1.8 : Mon Nov 07 2005 - 08:01:02 EST