At 08:59 PM 6/22/99 -0500, Glenn wrote:
>Brian Harper wrote:
>>>I think the confusion here results from the
>common association of fine-tuning with probability arguments.
>But fine-tuning itself does not rely on probability calculations.
>So, I think its best to say that the many worlds hypothesis provides
>a possible explanation for fine-tuning.<<
>
>I have heard this argument before but have found it less than
>convincing. The AP arguments, it seems to me, must be based upon an
>implicit probability argument. Take the nuclear resonance found by
>Hoyle which allowed carbon to be produced in stars. Why is this
>resonance note-worthy? Because of all the other possible configurations
>that COULD HAVE happened. Thus the fact that the precise value required
>for life is what we find in our universe is considered evidence of a
>finely tuned universe. The choice of one out of an innumerable plethora
>of hostile possibilities is what makes the anthropic principle work.
>
>Any comments?
>
Well, I obviously disagree. Unfortunately, one seldom finds a
precise definition of fine-tuning in the literature and one
has to read between the lines. For example, the Sciama paper
I quoted from would be utter nonsense if fine-tuning were
tied to probability calculations.
Briefly looking through my papers on the AP, the closest thing
I found to a definition of fine-tuning is the following from
Leslie
#"Recently, many have argued that either reality as a whole,
#or else the spatiotemporal region which we can see, is
#"fine-tuned" to life's needs, by which they mean that tiny
#changes in its basic properties would have excluded life
#forms of any kind. (Talk of "fine-tuning" does not presuppose
#a divine fine-tuner.)"
# -- John Leslie "Introduction" to <Physical Cosmology and
#Philosophy>, edited by Leslie, Macmillan, New York, 1990.
One of the things I do for a living is to try to develop
models. Leslie's definition of fine-tuning fits
perfectly with the way I would use the term in my own
work. Suppose I have some model Y which I want to produce
some response X (X in my case would usually be to fit
some data sets X1, X2, X3 etc. obtained in various special
cases). The model has several parameters a,b,c,d....
Parameter c, say, would be finely tuned wrt X if it must
fall in some very narrow range in order for Y-->X. I
would say c is finely tuned regardless of how it is obtained.
Probability doesn't have anything to do with it, unless
I were to make some bold claim that I had determined c
by picking numbers at random ;-).
Now let me go back to your final statement:
"The choice of one out of an innumerable plethora
of hostile possibilities is what makes the anthropic
principle work." -- Glenn
Not really. The main point I believe is the fine-tuning
as defined above combined with the appearance that the
laws of physics care about whether there is or isn't
life. Recall Sciama's comment:
"These finely tuned properties will probably also eventually
be accounted for by fundamental theory. But why should fundamental
theory _happen_ to lead to these properties?" -- D. Sciama
Brian Harper
Associate Professor
Applied Mechanics
The Ohio State University
"All kinds of private metaphysics and theology have
grown like weeds in the garden of thermodynamics"
-- E. H. Hiebert