author : Lee Smolin

In this gem of a book, Smolin describes his quest for a theory of the universe, a cosmological theory that explains why the universe is necessarily large and complex and the way it is. The old Newtonian, clockwork model of the universe is a picture of a cold, inhospitable, essentially sterile place in which life is a fluke, a vastly improbable statistical fluctuation, and in which all is futile, heading inexorably to the inevitable Heat Death. Fortunately it is a false picture, based on a 19th century understanding of physics. Unfortunately, it is also the picture that many believe to be what physics is still telling us. Smolin overthrows the old, showing how the 20th century physical theories of general relativity and quantum mechanics lead to a picture of a vast, vibrant, complex self-organising universe, hospitable to life, growing, and exhibiting ever more variety.

He takes a philosophical approach to describing physical theories. He points out that two of the mainstays of classical physics, reductionism and atomism, are simply incompatible: reductionism works by describing the whole in terms of its parts, whereas the fundamental "atoms" have no parts. Instead, he takes Leibniz' ideas of relative descriptions, and the principle of sufficient reason, very seriously. He shows how general relativity has moved us away from a theory with absolute space and time (implying some absolute, externally-imposed frame of reference) to a description based on the relationships between things, how position, motion, acceleration are relative concepts (hence the name, "relativity"). And he shows how quantum mechanics has moved us away from a theory of individual things acting independently, to one of entangled things all dependent on each other. And the consequences of these two theories seem to point to a universe that is necessarily large and complex.

Smolin starts by showing that the universe we live in appears to be vastly improbable. In particular, because it contains stars. Stars are necessary for life: they synthesise the sufficient variety of chemical elements (and particularly carbon) needed to build sufficiently complex systems, and they provide the long-term thermodynamic gradients needed for stable far-from-equilibrium systems. A universe full of stars is necessarily quite complex (for example, such a universe needs to contain carbon, to cool the interstellar medium sufficiently that stars can be made). Yet the fundamental physical constants need to be tuned astonishingly precisely to allow a universe with stars, and hence with life. Change those constants only very slightly, and there are no stars, and hence no life.

Why do those constants have the very finely-tuned values they do? Coincidence? Smolin says not, and provides a mechanism for giving them the values they have. General relativity predicts singularities; but we don't know, by definition, what happens at a singularity. Quantum mechanics suggests that maybe there was no singularity at the Big Bang, that maybe there are no singularities inside black holes. What if, Smolin speculates, inside black holes, the "singularity" actually produces a whole new universe? And what if the laws of physics (ie the values of the fundamental constants) are slightly different from those in the "parent" universe? Then each new "singularity" (be it a black hole, or the Big Crunch at the end of the universe) produces a universe slightly different from its parent. A form of natural selection can act: the random walk through value space eventually finds universes where black holes are abundant, and such universe generate many more child universes, with similar values of the constants, and hence will come to dominate. Any universe chosen at random will tend to be one that generates many black holes. Such universes are necessarily rather complex, and so are also good for life. A universe where black holes are abundant must have stars (to turn into black holes) and carbon (to help make stars). This is not simply wild speculation. It is proper scientific theory and is testable/falsifiable: it makes predictions that any universe with different values of the fundamental constants has fewer black holes.

This idea explains why the universe we live in is hospitable to life, without having to invoke special pleading such as the anthropic principle. But we have to give up some "neatness" if this idea is correct. If our universe is a result of natural selection by random walk over the space of values of the fundamental constants, these values may not be "simple" -- there may be no reason for the lower order digits to have the precise values they do (a value very close, but differing in all later digits, might also be tuned sufficiently well). [Is there any connection with Chaitin's work on algorithmic information theory here? Might some of the novelty come from the massive amount of information in the infinitely precise, but essentially random, constants? Or are they actually not infinitely precise, given the practical limitations of measuring them with a finite universe?]

Smolin moves on to discuss the consequences of GR and QM in some detail. He explains how GR gradually helped break the ideas of absolute space and time (it took some time, but GR broke free from coordinate systems approaches with the geometrical approach described by Misner, Thorne and Wheeler). He also explains how QM gradually helped break the idea of it being legitimate to talk about single isolated particles. (Again, this did not happen immediately, because early QM tackled descriptions of single particles. It wasn't until multi-particle systems were tackled that quantum entanglement was discovered and fully appreciated.) In order for these theories to make sense, the universe must be sufficiently complex (for example, if every thing is relative, the universe must be complex enough, have enough variety, that the things in it can be distinguished from each other purely in terms of their relationships with each other, not in terms of some non-existent absolute position).

This all leads on to some very deep stuff. In particular, Smolin points out that GR and QM are at best only partial cosmological theories, and so more work is needed. He speculates on what a full cosmological theory might be. And on the way, he packs in numerous little gems (such as the Bekenstein bound: a black hole has an entropy proportional to its surface area; no region of space can have more information than can the largest black hole that can occupy that space; so the maximum amount of information in a region of space grows as the surface area, and not the volume, of that space.) No review could possibly hope to cover the breadth and depth of the topics Smolin explores (for example: his explanation of how Leibniz' philosophy and gauge theories are linked, and his description of the process of continual star production in the spiral arms of galaxies, are a joy to read).

Smolin is careful to distinguish between accepted physical theory, theories that are still currently under development, and more speculative ideas. Some of the more interesting points are the more speculative ones! Nevertheless, this beautifully written and fascinating account, by a leading quantum gravity researcher, is of a universe vastly different, and vastly more interesting, than the one we are used to reading about in more "popular" accounts of physics.

Smolin sets out to explain three different routes to a theory of quantum gravity, and how they might all be leading to the same place, and does so brilliantly. The three roads are black hole thermodynamics, loop quantum gravity, and string theory.

He starts with a description of cosmological logic -- that we can only know about events in our past light cones, that different observers have different light cones (so know different things), and these light cones are growing (so we continually know more). Putting together a logic that allows reasoning under these conditions results in a very different world from that painted by an all-knowing Platonic style of logic. But this less absolute logic isn't some wooly NewAge "everything is relative, and so everything is true" idea -- if observers have overlapping light cones, they will agree on what they can deduce about the shared region. It transpires that there is relationship between this kind of logic and something the mathematicians have already come up with: a very hairy branch of Category Theory known as Topos Theory. (Now I think I understand why Kauffman makes a passing reference to Category Theory in his Investigations -- given he has worked with Smolin, he could be referring to these ideas.)

Smolin also discusses in detail the ideas of "background independent" theories -- ones where there is no framework of absolute time and space for particles to move against, but rather ones where space and time themselves are integral, evolving, changing parts of the cosmos, and in which there are no static things, only dynamic processes. He touched on this to some degree in his earlier marvelous book, The Life of the Cosmos, but goes into more detail and explanation here. And it is all explained so clearly -- I was particularly enthralled with the description of the relativity principle link between the well-known radiation emitted at a black hole's event horizon, and the weird Unruh radiation seen by an accelerating observer in empty space.

Loop quantum gravity is a theory about how space and time are constructed, but the resulting universe doesn't appear to have much else in it. String theory, on the other hand, has particles, but it is a background dependent theory. Work is beginning on how these might be two components of the final theory. Black hole thermodynamics links everything together, and seems to be providing a clue about the relationship between thermodynamic entropy and information. We end up with the weirdness of the Bekenstein bound (again, touched on in TLotC, but gone into in more depth here) and the holographic principle.

Interwoven with all this hard science are some great little vignettes -- Smolin is one of the key players in the loop quantum gravity strand. These serve to enliven and enrich the scientific ideas, and illuminate the scientific process, rather than detracting from them with arbitrary personal details, as happens all too often in the less well written popular science accounts. (It is interesting that scientist popularisers write about science, whereas journalist popularisers tend to write about the people. My interests lie with the science -- I can get people stories anywhere.)

There are some incredibly deep ideas here, explained brilliantly. All in all, this is a marvellous book. It contrasts nicely with TLotC, complementing that one's grand scope with some really fundamental hard science, told in a refreshing and comprehensible manner.