麻豆传媒

THE UNIVERSE is the way it is because if it weren鈥檛 we wouldn鈥檛 be here to
see it. Profound insight or empty truism? This is the 鈥渁nthropic principle鈥�, and
in the past few years it has been embraced by many cosmologists to explain some
of the most mystifying features of the Universe. But it leaves other scientists
feeling deeply troubled.

One of them is cosmologist Paul Steinhardt of Princeton University in New
Jersey, who claims that the anthropic principle is sloppy and unscientific.
鈥淚t鈥檚 corrupting science,鈥� he says. His view is shared by physicist Gordon Kane
of the University of Michigan in Ann Arbor, who believes that string theory will
make the anthropic principle redundant. The pro- and anti-anthropic camps are
debating just about the most profound question there is: why are we here? But
the final answer might not please either side.

What makes the debate more difficult is that a precise definition of the
anthropic principle is hard to come by. Roughly speaking, it asserts that our
existence restricts the possible values of any physical constants, because
simply to be observed, the Universe must allow life to exist. If we assume that
life anywhere in the Universe must be broadly like that on Earth, for example,
the Universe must allow stars to form. And the Universe must be old enough for
those stars to have created plenty of heavy elements鈥攖he building blocks
of biological molecules鈥攊n order to allow life to evolve.

Arguably, British astronomer Fred Hoyle was the first to evoke the anthropic
principle. In 1952, he used the fact that carbon exists to predict that
carbon-12 had an 鈥渆xcited鈥� energy state unknown at the time. If the state did
not exist, Hoyle reasoned, nuclear reactions inside stars could not assemble
carbon and all heavier elements from lighter nuclei, and so carbon-based life
such as ours would not exist.

The principle has gained popularity because of the apparent fine-tuning of
physics. For example, if the strong nuclear force were just a few per cent
stronger, the Sun would burn all its hydrogen fuel in less than a
second鈥攏ot long for intelligent life to evolve. 鈥淢any instances have been
found in which if a certain fundamental force were slightly weaker or stronger,
or if a certain fundamental particle were slightly lighter or heavier, there
would be no galaxies or stars or planets, and hence no human beings,鈥� says Max
Tegmark of the University of Pennsylvania. Martin Rees of the University of
Cambridge agrees: 鈥淚 think the fine-tuning `coincidences鈥� need some kind of
别虫辫濒补苍补迟颈辞苍.鈥�

鈥淓ither God fine-tuned the Universe for us to be here,鈥� says Tegmark, 鈥渙r
there are many universes, each with different values of the fundamental
constants, and not surprisingly we find ourselves in one in which the constants
have the right values to permit galaxies, stars and life.鈥�

Some have dubbed this the 鈥渕ultiverse鈥�. Inflation, one theory that aims to
describe the first split second of existence after the big bang, hints that the
observable Universe may simply be one bubble among an infinity of others in a
vast ocean of space. Conceivably, the other universes might be distant patches
of space where the fundamental constants are different.

Tegmark believes that the anthropic principle is becoming more widely
accepted. 鈥淭wo years ago, at a conference at Fermilab there was an audible hiss
when someone mentioned the A-word,鈥� he says. 鈥淏ut a couple of months ago at
another conference in Santa Monica, Steven Weinberg argued that the anthropic
principle may be the best explanation for the cosmological constant.鈥�

The cosmological constant, which is a measure of the curious repulsive force
exerted by empty space, is an embarrassing 10¹²³ times smaller than that
predicted by quantum theory. Weinberg explains this by assuming that the
constant takes on all possible values in all possible universes. Most universes
would accelerate madly (where the constant is large) or make nothing but black
holes (where it is large and negative). Only in a few freak universes where the
constant is tiny would galaxies, stars and planets arise.

So why does Steinhardt so vehemently oppose the anthropic principle? 鈥淚 have
several reasons,鈥� he says. 鈥淔irst, the anthropic principle is not testable,
which means it is not science.鈥�

According to Steinhardt, a statement has scientific meaning only if it can be
tested by an experiment or observation. He maintains that there is no such test
for the anthropic principle. 鈥淭his makes it entirely different from, say,
Newton鈥檚 general principle that the laws of physics are deterministic,鈥� he says.
鈥淭his was tested and found to be false in the realm of the atom.鈥�

Tegmark, however, argues that the anthropic principle is no more
sophisticated or controversial than the principles of logic that underpin a
statement like 1 + 1 = 2. 鈥淚f I have two competing theories, the anthropic
principle merely selects the theory which gives the greatest likelihood of
seeing what I see, given that I am here,鈥� he says. 鈥淭he anthropic principle is
no more than an application of probability theory. Unfortunately, the word
`principle鈥� suggests there is something deep about it. In fact, there鈥檚 nothing
to test.鈥�

But Steinhardt has a more telling criticism. One of the characteristics of
science, he says, is that you begin with a little and get out a lot. It鈥檚 an
efficient way of gaining knowledge. 鈥淲hat does the anthropic argument get you?
I鈥檓 not sure there鈥檚 enough to fill the back of a postage stamp.鈥�

Take the cosmological constant. The anthropic principle seems to provide a
rational explanation for why this constant has the value it has鈥攂ut does
that really tell us anything? Isn鈥檛 it just a way of feeling comfortable with
the way the Universe is? Contrast this with Newton鈥檚 law of universal
gravitation. Newton developed his theory to explain the orbits of the planets in
our Solar System, but it has many other observable consequences鈥攊t
predicts the existence of lunar tides and the movements of comets, moons,
asteroids and other stars.

So is the anthropic principle a useless but essentially harmless idea?
Steinhardt thinks it鈥檚 worse than that. He says it 鈥渄ulls sharp problems with an
air of explanation鈥濃�� it stops people from struggling to find really
fundamental solutions. He sees its increasing popularity as partly due to
impatience. 鈥淧eople want to know all the answers right now, but we have to
accept that this is not possible,鈥� he says. 鈥淚鈥檇 rather say `I don鈥檛 know鈥�.鈥�

Another serious objection to the anthropic principle is that 鈥渢here are no
rules of the game鈥�, says Steinhardt. He points out that there is no
justification for the belief that physical constants vary from one universe to
another. The problem, he says, is that there is no underlying 鈥渕etatheory鈥� which
predicts how they vary.

We know that some parameters, such as the distance between the Earth and the
Sun, can vary. Before Newton, there had been high hopes that this distance would
turn out to be a fundamental constant, perhaps fixed by the geometry of cubes
and other solid figures. However, Newtonian dynamics allowed a whole continuum
of Sun-Earth distances. This is a little like the sort of metatheory that
Steinhardt is talking about. For life to exist on Earth, the Earth must be
neither too cold nor too hot鈥攕o we shouldn鈥檛 be surprised that the
Earth-Sun distance is just right for water to flow on the Earth鈥檚 surface.

But even this application of the principle is flawed. 鈥淓ven now we cannot say
how common planets are, or whether planets need to be like ours to foster life,鈥�
says Steinhardt. For instance, if silicon-based machines could evolve, they
might find themselves on a planet closer than Mercury is to the Sun. If it
turned out that life could exist in a huge range of physical environments, we
would find ourselves in the special position of being cool, watery, carbon-based
observers. The anthropic principle would lose its power.

But in the case of the cosmological constant, says Steinhardt, there isn鈥檛
even a theory predicting that it can take on a range of values. He maintains
that such a theory, like Newtonian dynamics, would be bound to have other
consequences鈥攑erhaps in the microscopic world鈥攚hich could be
measured. 鈥淚f there鈥檚 a breakthrough and the anthropic people can come up with
an abundantly predictive, testable metatheory, I would be happy to accept
颈迟.鈥�

Rees believes that such a metatheory could come from Andrei Linde鈥檚 idea of
eternal inflation. In that model, space spontaneously 鈥渋nflates鈥� to give birth
to new universes, which give birth to yet more universes, and so on, ad
infinitum. 鈥淭here are two key questions for 21st-century physics/cosmology,鈥� he
says. 鈥淲hen we understand the physics of the inflationary era, will it indeed
predict multiple big bangs, as in the simulations of Linde and his colleagues?
And do the underlying physical laws allow the different big bangs to end up
governed by different low-energy physics?鈥� According to Rees, if the answer
to both these questions is 鈥測es鈥�, then anthropic selection would explain the
apparent fine-tuning in our Universe.

Tegmark would like to see this fine-tuning quantified. He believes that
scientists have not performed such calculations because of
鈥渁nthropophobia鈥濃�攖he principle makes them uneasy. The build-up of heavy
elements inside stars, for example, depends on the electromagnetic and strong
coupling constants, which determine the strengths of the electromagnetic and
strong nuclear forces. 鈥淚f someone calculated the detailed consequences for
stars of varying these parameters, we could find out how small is the island in
parameter space we live on,鈥� he says. That would tell us just how finely tuned
the Universe is. Rees adds that if we found ourselves living on a highly
improbable part of this island, that would invalidate the anthropic
principle.

Useless speculation

All this may be useless speculation, however, because there are reasons to
believe the constants can鈥檛 vary. 鈥淚 would say the examples of fine-tuning
people have found aren鈥檛 really there,鈥� says Steinhardt. He cites Grand Unified
Theories (GUTs), which attempt to unify the three non-gravitational forces
of nature. In GUTs, the strengths of the electromagnetic and strong nuclear
forces are intimately linked. 鈥淚f you vary one, the other changes too,鈥� says
Steinhardt. So the argument that all the constants must be individually
fine-tuned begins to look shaky. Kane agrees: 鈥淚t鈥檚 simply not so that
increasing the strong force or decreasing the electromagnetic force affects how
the Sun works,鈥� he says. 鈥淢ost of the old arguments are just misleading
苍耻尘别谤辞濒辞驳测.鈥�

Kane believes that string theory provides even more powerful arguments
against fine-tuning. In string theory, the fundamental entities of reality are
tiny strings vibrating in 9-dimensional space, and all the constants of nature
depend on a single fundamental parameter. 鈥淪o if the theory is right there will
be no room to vary any of the constants the anthropic people like to vary,鈥� says
Kane. He and his colleagues, Malcolm Perry and Anna Zytkow of the University of
Cambridge, have just written a paper outlining this argument, titled 鈥淭he
beginning of the end of the anthropic principle鈥�.

But string theory does allow the vacuum to adopt a range of different states
in different universes, which should, for example, have different cosmological
constants. 鈥淚n a sense, the ensemble of universes is replaced by something with
a better theoretical basis, the many vacua of string theory,鈥� admits Kane.
鈥淒ifferent vacua will lead to different universes, and only some will be right
for life to emerge.鈥�

So have string theorists, in rejecting the anthropic principle for a final
theory, let it in by the back door? Is string theory the metatheory that
Steinhardt is looking for? 鈥淪ure, superstring may some day prove to be a
metatheory with lots of predictive power,鈥� says Steinhardt. 鈥淏ut I don鈥檛 think
Newton鈥檚 theory or string theory include the anthropic principle鈥擨 think
they replace 颈迟.鈥�

Kane is also dubious. 鈥淯ntil the vacuum structure of string theory is
understood better, it looks a lot like the multiple universes, but not the
same,鈥� he says. That鈥檚 because the various regions might be causally connected,
affecting each other鈥檚 states.

For his part, Tegmark is sceptical of string theory. 鈥淚t is emerging as the
modern version of the emperor with no clothes,鈥� he says. 鈥淪o far, it has
predicted the value of none of the physical constants, and I鈥檓 willing to bet
that numbers like the electromagnetic coupling constant will never be derivable
from pure math.鈥� 鈥淚s he serious?鈥� responds Steinhardt. 鈥淲here can we line up to
take that bet?鈥�

But if Steinhardt wins the bet, the consequences may be uncomfortable. The
theorists hope that string theory will be mathematically inevitable鈥攖he
only logically consistent theory of the Universe. But if so, and if string
theory pins down every physical constant, then the fine-tuning for life will
turn out to be hard-wired into mathematics. 鈥淚n that case, string theory will be
a great argument for design,鈥� says Tegmark. He is reminded of Carl Sagan鈥檚
novel, Contact, in which mathematicians calculate &pgr; to billions of
decimal places and suddenly find the digits getting non-random. 鈥淚t turns
out there is a message written in a fundamental constant of mathematics鈥攁
message from the Creator,鈥� says Tegmark. Likewise, if the mathematics makes life
inevitable, people might start using string theory as an argument for the
existence of God.

String theorists believe they are following a path to an ultimate rational
theory that is a cut above the anthropic alternative. They might be upset, to
say the least, to find God in their equations.