Inspired by this paper a few of us at UCL have begun thinking through some of the problems. I’ll try and summarize things here as I think that attempting to answer some of these fundamental questions provides a nice illustration of what we do and do not know.

The first problem George Ellis brings up is the simple fact of the uniqueness of the Universe. This has profound consequences; cosmology on the largest scale is not an experimental science (we can’t rerun the history of the Universe and see what happens) and it is not even a comparable one in the way that, say, galaxy formation is. If you want to know how galaxies form you can compare many different galaxies, each of which evolved more or less independently.

This means that it’s impossible to establish large-scale laws of physics for the Universe. For example, it is a fundamental assumption of modern cosmology that the Universe is homogeneous, with material smoothly distributed on the largest scales. Although we can imagine testing the extent to which this is true, we will never be able to establish whether this had to have been so, or whether it’s just a cosmic coincidence.

This issue comes up a lot already. The results from WMAP’s studies of the cosmic microwave background indicate that there is less power than expected on large scales. Is this an example of ‘cosmic variance’ – our Universe just happening to be different from the ‘average’ on these large scales?

But there’s another problem. I’ve talked about our Universe being ‘average’ but from our perspective within it it is the only one we can ever experience, see or test. How, then, do we define average?

This sort of reasoning does not threaten the conclusions of the currently standard cosmological model, but it does challenge us to think carefully about the limits on our knowledge. There will be much more to come over the next few months on this subject, so please comment and let me know if you’re interested (or not!).