On Mon, 6 Oct 2003, Ioan Oprea wrote:

The point I was trying to make is that the anthropic principle is not
something that results from the theory (neither string theory nor any
other one). And the fact that we have to use it reflects rather an
inability of our theory than its success.

I totally agree. Every time physicists (and one might even say "pure
scientists") face some truly difficult problems, some of them start to
give up and accept not-quite-scientific explanations - especially the
anthropic principle - as the last chance. The appearance of the anthopic
principle is therefore a symptom of the fact that the physicists have a
problem to explain a known fact scientifically.

In the case of the recent popularity of the anthropic principle among some
string theorists, the truly difficult problem was the cosmological
constant problem. The problem "doubled" once the astrophysical
observations showed that not only that Lambda is much much smaller than
any reasonable scale that one can imagine to emerge from a high energy
theory, but it is - despite its tiny value - nonzero. We - I mean all
theoretical and particle physicists - just don't have a natural
explanation of this fact, and therefore physicists - inspired by Bousso
and Polchinski - started to propose scenarios meant to generate a huge
number of metastable vacua in the theory which has the virtue that one of
them is more likely to have the right, minuscule value of the cosmological
constant. Some people started to like these highly-degenerate vacua so
much that they declared that string theory must certainly predict this
huge number of vacua, and therefore the anthropic considerations will be
always necessary in string theory.

In my opinion, this whole direction of research might be, and it's quite
likely, on a wrong track. We should still try to find a formalism that
allows to compute the cosmological constant and other SUSY-breaking
effects reliably and scientifically, and the small value will emerge as a
consequence of a pretty general, rigid and rather unique small set of
vacua - such as the heterotic strings on Calabi-Yau 3-folds. This will
make these huge, artificially created multiverses of stringy vacua
irrelevant for physics, and once we understand the theory really well, we
might be able to prove that there's something wrong with this "huge
landscape".

For example, you could ask that the solution
describe a universe with a certain amount of Fe atoms (I dont't know
the proportion of Fe in our universe, but you could adjust the
percentage in order to include our universe among the selected
solutions). Why not use this to select the right solution?

Right! This is very similar to my favorite example. Before physicists
understood Schrodinger's equation (that's responsible for the properties
of all the atoms), they could have argued that the world has hundreds of
elementary particles (the atoms or - at least - their nuclei) whose
properties (masses, spectral lines etc.) must be chosen exactly as they
are observed, otherwise the world would not admit life. Well, today we can
calculate all the properties of these atoms from a handful or parameters,
and therefore we know that such an anthropic proposal would have been
silly. Unfortunately we can't calculate the cosmological constant after
SUSY breaking reliably today, and therefore I can't show whether/that the
case of the cosmological constant is analogous. Yes, I think it may be,
and a physicist who wants to continue to try to make nontrivial progress
in this direction *should* believe that these two examples are analogous.

The choice of the constraint on the solutions is (almost) completely
random. And I might be missing something here, but this doesn't look
like a TOE, but resembles more a creation myth. Each theorist could
have his own favourite criterion of selecting the right solutions, the
only requirement being that our universe is included.

Exactly. This is not real science. Some advocates of the anthropic
principle in string theory argue that string theory could still explain
the quantum properties of black holes. Which black holes? should we ask.
The properties of the observed black holes? Why should we believe that the
theoretical ones have anything to do with the observed ones, unless we can
calculate and compare at least one number? Once we admit that we don't
know physics at higher energies (the particle spectrum etc.), we can't
claim anything specific about quantum gravity either. If we were satisfied
with the claims that string theory describes the black hole entropy
consistently and this is why string theory is better, theoretical physics
- and string theory in particular - would become a religion.

The internal mathematical consistency is what makes string theory
attractive, probable and promising - but it does not prove it!

The only real reason why some people are happy to accept such a scenario
is that we have not quite understood one number - the cosmological
constant. In the light of string theory's shocking ability to give us -
naturally - all the required ingredients for particle physics and gravity,
such a surrender does not look reasonable to me, especially because it
leads us to look at complicated vacua that don't explain the observed
facts too naturally. The old-fashioned vacua - like heterotic strings on
Calabi-Yau's - and some newer ones predict (very naturally) the particle
spectrum; the gauge groups etc. We don't quite know how to calculate
quantities after SUSY breaking, and Lambda is no exception. Why should we
suddenly transfer our belief to some other models that don't give too
convincing picture of reality and that look artificially, just because of
the single number?

__________________________________________________ ____________________________
E-mail: fax: +1-617/496-0110 Web: http://lumo.matfyz.cz/
phone: work: +1-617/496-8199 home: +1-617/868-4487
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Superstring/M-theory is the language in which God wrote the world.