Maybe the psychological reason for this amazing blindness has something to do with the fact that many people have not had their consciousness raised, as biologists have, by natural selection and its power to tame improbability. J. Anderson Thomson, from his perspective as an evolutionary psychiatrist, points me to an additional reason, the psychological bias that we all have towards personifying inanimate objects as agents. As Thomson says, we are more inclined to mistake a shadow for a burglar than a burglar for a shadow. A false positive might be a waste of time. A false negative could be fatal. In a letter to me, he suggested that, in our ancestral past, our greatest challenge in our environment came from each other. 'The legacy of that is the default assumption, often fear, of human intention. We have a great deal of difficulty seeing anything other than human causation.' We naturally generalized that to divine intention. I shall return to the seductiveness of 'agents' in Chapter 5.

Biologists, with their raised consciousness of the power of natural selection to explain the rise of improbable things, are unlikely to be satisfied with any theory that evades the problem of improbability altogether. And the theistic response to the riddle of improbability is an evasion of stupendous proportions. It is more than a restatement of the problem, it is a grotesque amplification of it. Let's turn, then, to the anthropic alternative. The anthropic answer, in its most general form, is that we could only be discussing the question in the kind of universe that was capable of producing us. Our existence therefore determines that the fundamental constants of physics had to be in their respective Goldilocks zones. Different physicists espouse different kinds of anthropic solutions to the riddle of our existence.

Hard-nosed physicists say that the six knobs were never free to vary in the first place. When we finally reach the long-hoped-for Theory of Everything, we shall see that the six key numbers depend upon each other, or on something else as yet unknown, in ways that we today cannot imagine. The six numbers may turn out to be no freer to vary than is the ratio of a circle's circumference to its diameter. It will turn out that there is only one way for a universe to be. Far from God being needed to twiddle six knobs, there are no knobs to twiddle.

Other physicists (Martin Rees himself would be an example) find this unsatisfying, and I think I agree with them. It is indeed perfectly plausible that there is only one way for a universe to be. But why did that one way have to be such a set-up for our eventual evolution? Why did it have to be the kind of universe which seems almost as if, in the words of the theoretical physicist Freeman Dyson, it 'must have known we were coming'? The philosopher John Leslie uses the analogy of a man sentenced to death by firing squad. It is just possible that all ten men of the firing squad will miss their victim. With hindsight, the survivor who finds himself in a position to reflect upon his luck can cheerfully say, 'Well, obviously they all missed, or I wouldn't be here thinking about it.' But he could still, forgivably, wonder why they all missed, and toy with the hypothesis that they were bribed, or drunk.

This objection can be answered by the suggestion, which Martin Rees himself supports, that there are many universes, co-existing like bubbles of foam, in a 'multiverse' (or 'megaverse', as Leonard Susskind prefers to call it).*) The laws and constants of any one universe, such as our observable universe, are by-laws. The multi- verse as a whole has a plethora of alternative sets of by-laws. The anthropic principle kicks in to explain that we have to be in one of those universes (presumably a minority) whose by-laws happened to be propitious to our eventual evolution and hence contemplation of the problem.

An intriguing version of the multiverse theory arises out of considerations of the ultimate fate of our universe. Depending upon the values of numbers such as Martin Rees's six constants, our universe may be destined to expand indefinitely, or it may stabilize at an equilibrium, or the expansion may reverse itself and go into contraction, culminating in the so-called 'big crunch'. Some big crunch models have the universe then bouncing back into expansion, and so on indefinitely with, say, a 20-billion-year cycle time. The standard model of our universe says that time itself began in the big bang, along with space, some 13 billion years ago. The serial big crunch model would amend that statement: our time and space did indeed begin in our big bang, but this was just the latest in a long series of big bangs, each one initiated by the big crunch that terminated the previous universe in the series. Nobody understands what goes on in singularities such as the big bang, so it is conceivable that the laws and constants are reset to new values, each time. If bang-expansion-contraction-crunch cycles have been going on for ever like a cosmic accordion, we have a serial, rather than a parallel, version of the multiverse. Once again, the anthropic principle does its explanatory duty. Of all the universes in the series, only a minority have their 'dials' tuned to biogenic conditions. And, of course, the present universe has to be one of that minority, because we are in it. As it turns out, this serial version of the multiverse must now be judged less likely than it once was, because recent evidence is starting to steer us away from the big crunch model. It now looks as though our own universe is destined to expand for ever.

Another theoretical physicist, Lee Smolin, has developed a tantalizingly Darwinian variant on the multiverse theory, including both serial and parallel elements. Smolin's idea, expounded in The Life of the Cosmos, hinges on the theory that daughter universes are born of parent universes, not in a fully fledged big crunch but more locally in black holes. Smolin adds a form of heredity: the fundamental constants of a daughter universe are slightly 'mutated' versions of the constants of its parent. Heredity is the essential ingredient of Darwinian natural selection, and the rest of Smolin's theory follows naturally. Those universes that have what it takes to 'survive' and 'reproduce' come to predominate in the multiverse. 'What it takes' includes lasting long enough to 'reproduce'. Because the act of reproduction takes place in black holes, successful universes must have what it takes to make black holes. This ability entails various other properties. For example, the tendency for matter to condense into clouds and then stars is a prerequisite to making black holes. Stars also, as we have seen, are the precursors to the development of interesting chemistry, and hence life. So, Smolin suggests, there has been a Darwinian natural selection of universes in the multiverse, directly favouring the evolution of black hole fecundity and indirectly favouring the production of life. Not all physicists are enthusiastic about Smolin's idea, although the Nobel Prize-winning physicist Murray Gell-Mann is quoted as saying: 'Smolin? Is he that young guy with those crazy ideas? He may not be wrong.'⁷⁰ A mischievous biologist might wonder whether some other physicists are in need of Darwinian consciousness- raising.

It is tempting to think (and many have succumbed) that to postulate a plethora of universes is a profligate luxury which should not be allowed. If we are going to permit the extravagance of a multiverse, so the argument runs, we might as well be hung for a sheep as a lamb and allow a God. Aren't they both equally un- parsimonious ad hoc hypotheses, and equally unsatisfactory? People who think that have not had their consciousness raised by natural selection. The key difference between the genuinely extravagant God hypothesis and the apparently extravagant multi- verse hypothesis is one of statistical improbability. The multiverse, for all that it is extravagant, is simple. God, or any intelligent, decision-taking, calculating agent, would have to be highly improbable in the very same statistical sense as the entities he is supposed to explain. The multiverse may seem extravagant in sheer number of universes. But if each one of those universes is simple in its fundamental laws, we are still not postulating anything highly improbable. The very opposite has to be said of any kind of intelligence.