Re: Bad Bayesian - no biscuit! (was A New Year's gift for Bayesians)

From: Thomas Buckner (
Date: Fri Jan 21 2005 - 04:49:11 MST

--- Marc Geddes <> wrote:

> Uh...there's a bit of a puzzle here that I ran
> into
> when debating J.C.Wright.
> The trouble is that according to Bayes we can't
> actually assign a probability of 0 or unity to
> anything. And the only natural interpretation
> of the
> probabilities as I understood it is that they
> were
> referring to the frequency of space-time
> systems in
> the multi-verse. For instance a probability of
> 99%,
> meant that in 99 out of 100 alternative
> space-times
> such and such a thing was true.
> So are you sure it actually makes sense to
> assign a
> probability of say 99% that something is
> *absolutely*
> true? To say that something is truly
> 'Universal' is
> to say that there is no region of the Tegmark
> multiverse where the something is not true.
> But in
> that case, what would the counter-factual mean?
> How
> could the 1% probability of falsehood have any
> meaning?
Well, physical constants are absolutely true, in
our Hubble volume and within the time we have
been able to measure them. They are *locally
absolute* with the understanding that local means
anywhere we can go or see. Speculation about
these constants beyond our Hubble volume or in
other Tegmark levels is like counterfactual (what
if?) speculation among historians, a parlor game,
very entertaining but impossible to resolve.
Again, on this side of the line, *science*; on
that side of the line, *philosophy*.
Tom Buckner

Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:50 MDT