A new paper by Cambridge
physicist Stephen Hawking and Thomas Hertog of CERN
(firstname.lastname@example.org) suggests that it can.
The leading explanation
for the observed acceleration of the expansion of the universe is
that a substance, dark energy, fills the vacuum and produces a
uniform repulsive force between any two points in space -- a sort of
anti-gravity. Quantum field theory allows for the existence of such
a universal tendency. Unfortunately, its prediction for the value of
the density of dark energy (a parameter referred to as the
cosmological constant) is some 120 orders of magnitude larger than
the observed value.
In 2003, cosmologist Andrei Linde of Stanford
University and his collaborators showed that string theory allows
for the existence of dark energy, but without specifying the value
of the cosmological constant. String theory, they found, produces
a mathematical graph shaped like a mountainous landscape, where
altitude represents the value of the cosmological constant. After
the big bang, the value would settle on a low point somewhere
between the peaks and valleys of the landscape. But there could be
on the order of 10500 possible low points -- with different
corresponding values for the cosmological constant -- and no obvious
reason for the universe to pick the one we observe in nature.
Some experts hailed this multiplicity of values as a virtue of the
theory. For example, Stanford University's Leonard Susskind in his
book "The Cosmic Landscape: String Theory and the Illusion of
Intelligent Design," argues that different values of the cosmological
constant would be realized in different parallel worlds -- the pocket
universes of Linde's "eternal inflation" theory. We would just
happen to live in one where the value is very small. But critics
see the landscape as exemplifying the theory's inability to make
The Hawking/Hertog paper is meant to address this concern. It looks
at the universe as a quantum system in the framework of string
theory. Quantum theory calculates the odds a system will evolve a
certain way from given initial conditions, say, photons going
through a double slit and hitting a certain spot on the other side.
You repeat your experiment often enough and then you check that the
odds you predicted were the correct ones.
In Richard Feynman's
formulation of quantum theory, the probability that a photon ends up
at a particular spot is calculated by summing up over all possible
trajectories for the photon. A photon goes through multiple paths at
once and can even interfere with its other personas in the process.
Hawking and Hertog argue that the universe itself must also follow
different trajectories at once, evolving through many simultaneous,
parallel histories, or "branches." (These parallel universes are not
to be confused with those of eternal inflation, where multiple
universes coexist in a classical rather than in a quantum sense.)
What we see in the present would be a particular, more or less
probable, outcome of the "sum" over these histories. In particular,
the sum should include all possible initial conditions, with all
possible values of the cosmological constant.
But applying quantum theory to the entire universe -- where the
experimenters are part of the experiment -- is tricky. Here you have
no control over the initial conditions, nor can your repeat the
experiment again and again for statistical significance. Instead,
the Hawking-Hertog approach starts with the present and uses what we
know about our branch of the universe to trace its history
backwards. Again, there will be multiple possible branches in our
past, but most can be ignored in the Feynman summation because they
are just too different from the universe we know, so the probability
of going from one to the other is negligible.
For example, Hertog
says, knowledge that our universe is very close to being flat could
allow one to concentrate on a very small portion of the string
theory landscape whose values for the cosmological constant are
compatible with that flatness. That could in turn lead to
predictions that are experimentally testable. For example, one could
calculate whether our universe is likely to produce the microwave
background spectrum we actually observe.
Physical Review D, upcoming article
Contact Thomas Hertog, email@example.com