“That part is extremely simple. Unlike Henry Buzzo and Yasuko Nishide, I’m not looking for a way to make ‘our’ Big Bang seem like less of a ‘coincidence.’ Buzzo and Nishide both take the view that an infinite number of universes must have arisen out of pre-space – freezing out of that perfect symmetry with different sets of physical laws. And they both aim to reevaluate the probability of a universe ‘more-or-less like our own’ being included in that infinite set. It’s relatively easy to find a TOE in which our universe is possible, but freakishly unlikely. Buzzo and Nishide define a successful TOE as one which guarantees that there are so many universes similar to our own that we’re not unlikely at all – that we’re not some kind of miraculous, perfect bullseye on a meta-cosmic dartboard, but just one unexceptional point on a much larger target.”

I said, “A bit like proving – from basic astrophysical principles – that thousands of planets in the galaxy should have carbon-and-water-based life, and not just Earth.”

“Yes and no. Because … yes, the probability of other Earth-like planets can be computed from theory, alone – but it can also be validated by observation. We can observe billions of stars, we’ve already deduced the existence of a few thousand extrasolar planets – and eventually, we’ll visit some of them, and find other carbon-and-water-based life. But – although there are no end of elegant frameworks for assigning probabilities to hypothetical other universes … there is no prospect of observing or visiting them, no conceivable method for checking the theory. So I don’t believe we should choose a TOE on that basis.

“The whole point of moving beyond the Standard Unified Field Theory is that, one, it’s an ugly mess, and two, you have to feed ten completely arbitrary parameters into the equations to make them work. Melting total space into pre-space – moving to an All-Topologies Model – gets rid of the ugliness and the arbitrary nature of the SUFT. But following that step by tinkering with the way you integrate across all the topologies of pre-space – excluding certain topologies for no good reason, throwing out one measure and adopting a new one whenever you don’t like the answers you’re getting – seems like a retrograde step to me. And instead of ‘setting the dials’ of the SUFT machine to ten arbitrary numbers, you now have a sleek black box with no visible controls, apparently self-contained – but in reality, you’re just opening it up and tearing out every internal component which offends you, to much the same effect.”

“Okay. So how do you get around that?”