150 Cosmological models
came into being rather than any of the other options. Some particular solutions of
the equations have been implemented rather than the other possibilities; boundary
conditions choosing one set of solutions over others have still been set, even if
they are not technically initial conditions set at a finite time in the past.
3.9.1.3 Initial conditions are irrelevant because they all happened
The idea of an ensemble of universes, mentioned earlier, is one approach that
sidesteps the problem of choice of specific initial data, because by hypothesis
all that can occur has then occurred. Anthropic arguments select the particular
universe in which we live from all those in this vast family (see e.g. [57,70]). This
is again an intriguing and ingenious idea, extending to a vast scale the Feynman
approach to quantum theory. However, there are several problems.
First, it is not clear that the selection of universes from this vast family by
anthropic arguments will necessarily result in as large and as isotropic a universe
as we see today; here one runs up against the unsolved problem of justifying a
choice of probabilities in this family of universes. Second, this proposal suffers
from complete lack of verifiability. In my view, this means this is a metaphysical
rather than scientific proposal, because it is completely untestable. And in the
end, this suggestion does not solve the basic issue in any case, because then one
can ask:Why does this particular ensemble exist, rather than a different ensemble
with different properties?; and the whole series of fundamental questions arises
all over again, in an even more unverifiable form than before.
3.9.2 The explanation of initial conditions
The explanation of initial conditions has been the aim of the family of theories
one can label collectively asquantum cosmologyand the more recent studies of
string cosmology.
3.9.2.1 Explanation of initial conditions from a previous state of a different
nature
One option has been explaining the universe as we see it as arising from some
completely different initial state, for example:
- proposals forcreation of the universe as a bubble formed in a flat spacetime
or de Sitter spacetime, for example Tryon’s vacuum fluctuations and Gott’s
open bubble universes; or - Vilenkin’stunnelling universewhich arises from a state with no classical
analogue (described as ‘creation of the universe from nothing’, but this is
inaccurate).
These proposals (like the proposals by Hartle and Hawking, Hawking and
Turok, and Gott and Liu previously mentioned; for a comparative discussion and