Monday, November 24, 2008

Cosmogenic drift

According to current mathematical physics, there are many aspects of our physical universe which are contingent rather than necessary. These include such things as the values of the numerous free parameters in the standard model of particle physics, and the parameters which specify the initial conditions in general relativistic models of the universe. The values of these parameters cannot be theoretically derived, and need to be determined by experiment and observation.

Lee Smolin proposed that the values of these parameters can be explained by postulating that we live in a population of universes evolving by natural selection. He suggests that in those universes where black holes form, a child universe is created inside the event horizon of the black hole. Specifically, Smolin's proposal is that "quantum effects prevent the formation of singularities, at which time starts or stops. If this is true, then time does not end in the centers of black holes, but continues into some new region of space-time...Going back towards the alleged first moment of our universe, we find also that our Big Bang could just be the result of such a bounce in a black hole that formed in some other region of space and time." (The Life of the Cosmos, 1997, p93).

However, Smolin's scenario cannot explain why our universe is relativistic rather than non-relativistic, and it cannot explain why our universe is a quantum universe rather than a classical universe, because the occurrence of black holes requires a relativistic universe, and the occurrence of a 'bounce' inside the horizon of a black hole requires a quantum universe.

Smolin's hypothesis depends upon the assumption that there is a quantum relativistic universe at the outset. One can ask for an explanation of why there should be such a universe, rather than a universe in which, say, Newtonian gravity governs the large-scale structure of space-time, or in which classical mechanics and classical field theories govern the behaviour of any particles and fields which exist. The existence of a quantum relativistic universe seems to be contingent rather than necessary. There is, therefore, a need to explain the existence of a quantum relativistic universe.

My own proposal for such an explanation is to posit the existence of a random process, cosmogenic drift, which was responsible for the evolution of a quantum relativistic universe prior to the operation of cosmological natural selection.

In evolutionary biology it is known that evolution by natural selection is not the only important evolution process, and that in the absence of selection pressures, the evolution of a population will be dominated by random variations in the genome, a process called genetic drift. Similarly, the proposal made here suggests that the values of the parameters of physics cannot be wholly explained by cosmological evolution by natural selection. However, whilst genetic drift is a process which applies to a population of biological entities reproducing with inheritance and random mutation, the cosmological process postulated here is not restricted to reproducing entities, and in particular is postulated as a necessary prelude to the creation of a population of reproducing universes.

To best explain cosmogenic drift, we shall need a concept from evolutionary biology known as the fitness landscape. Each point on this landscape corresponds to a different combination of genes, and the height of the landscape at each point represents the average number of progeny, produced by an organism with that combination of genes, which themselves survive to reproduce. The height of the landscape therefore represents the 'fitness' of each possible genotype. Each progenitor produces offspring with genomes in a small neighbourhood of the progenitor's position in the landscape. In those parts of the landscape where selection pressures are weak, none of the progeny will have a greater fitness. When selection pressures are weak, the fitness landscape is therefore almost flat. Evolution of a biological population across a flat part of the fitness landscape will be driven by random diffusion. In contrast, in those parts of the landscape where selection operates, the landscape will possess gradient. In these parts of the landscape, some of the progeny produced within a small neighbourhood of one genotype will lie at a slightly greater height because they yield a greater number of progeny which themselves survive to reproduce. As a consequence, the population will come to be dominated by this new genotype, and will take a step-up to a slightly greater height in the fitness landscape. This is biological evolution by natural selection.

Smolin suggested that there is a cosmic fitness landscape analogous to the biological one, with each point corresponding to a combination of values for the parameters of physics, and the height at each point representing the number of progeny produced by a universe with that combination of parameters. I would like to extend this proposal by postulating that the lowest level of the cosmic fitness landscape is a flat region corresponding to all the possible types of universe which do not reproduce. Each point in the cosmic landscape here has a height of zero because none of these universes yield any progeny, and the landscape is flat because natural selection cannot operate in the absence of reproduction. Evolution does, nevertheless, occur in this part of the cosmic landscape. Universes, I propose, evolve by random diffusion in flat parts of the cosmic fitness landscape. In other words, universes which cannot reproduce evolve by cosmogenic drift. Eventually, however, a universe evolving by random cosmogenic drift will evolve into a quantum relativistic universe, a universe type capable of reproducing. The part of the cosmic fitness landscape containing universes capable of reproduction corresponds to a mountain in the landscape. Here there are gradients, and evolution by natural selection operates, as suggested by Smolin, in the same manner it operates in the biological fitness landscape.

2 comments:

The Dandy Highwayman said...

That's a rather nice idea. So, how do we tell if we're on the gradient of a local maximum or a global maximum?

Gordon McCabe said...

If you had a complete and final theory of everything, you could chart the cosmic fitness landscape, work out where we are in it, see if there's more than one local maximum, and then compare our local maximum to all the other maxima, local or global.