We come finally to the extreme edge of cosmological speculation: string cosmology. These models are based on an alternative to the standard quark model of elementary particle physics. So-called string theory (or M-theory) conceives of the fundamental building blocks of matter to be, not particles like quarks, but tiny vibrating strings of energy. String theory is so complicated and embryonic in its development that all its equations have not yet even been stated, much less solved. But that has not deterred some cosmologists from trying to craft cosmological models based on concepts of string theory to try to avert the beginning predicted by standard Big Bang cosmology.

The most celebrated of these scenarios in the popular press has been the so-called ekpyrotic scenario championed by Paul Steinhardt.[xvi] In the most recent revision, the cyclic ekpyrotic model, we are asked to envision two three-dimensional membranes (or ‘branes’ for short) existing in a five-dimensional space-time (Fig. 9). One of these branes is our universe. These two branes are said to be in an eternal cycle in which they approach each other, collide, and retreat again from each other. It is the collision of the other brane with ours that causes the expansion of our universe. With each collision, the expansion is renewed. Thus, even though our three-dimensional universe is expanding, it never had a beginning.

[…no picture supplied in original post…]

Now apart from its speculative nature the ekpyrotic scenario is plagued with problems.[xvii] For example, the Horava-Witten version of string theory on which the scenario is based requires that the brane on which we live have a positive tension. But in the ekpyrotic scenario it has a negative tension in contradiction to the theory. Attempts to rectify this have been unsuccessful. Second, the model requires an extraordinary amount of ad hoc fine turning. For example, the two branes have to be so perfectly aligned that even at a distance of 10³⁰ greater than the space between them, they cannot deviate from being parallel by more than 10^-60. There is no explanation at all for this extraordinary setup. Third, the collapsing and retreating branes are the equivalent of a 4-D universe which goes through an eternal cycle of contractions and expansions. In this sense, the cyclic ekpyrotic model is just the old oscillating model writ large in five dimensions. As such, it faces exactly the same problem as the original: there is no way for the universe to pass through a singularity at the end of each cycle to begin a new cycle and no physics to cause a non-singular bounce. Finally, even if the branes could bounce back, there is no means of the physical information in one cycle being carried through to the next cycle, so that the ekpyrotic scenario has been unable to deliver on its promises to explain the large-scale structure of the observable universe. These are just some of the problems afflicting the model. It is no wonder that Andrei Linde has recently complained that while the cyclic ekpyrotic scenario is ‘very popular among journalists,’ it has remained ‘rather unpopular among scientists’ (Linde 2002: 8).

But let all that pass. Perhaps all these problems can be somehow solved. The more important point is that it turns out that, like the chaotic inflationary model, the cyclic ekpyrotic scenario cannot be eternal in the past. In September of 2001 Borde and Vilenkin, in cooperation with Alan Guth, were able to generalize their earlier results on inflationary models in such a way to extend their conclusion to other models. Specifically, they note, ‘Our argument can be straightforwardly extended to cosmology in higher dimensions,’ specifically brane-cosmology.[xviii] According to Vilenkin, ‘It follows from our theorem that the cyclic universe is past-incomplete’,[xix] that is to say, the need for an initial singularity has not been eliminated. Therefore, such a universe cannot be past-eternal.