New physics from the CMB

2.3.2. Defects

Since I had many discussions at this meeting about alternatives to inflation, let me dwell a little on that subject here. The only real competitor to inflation has been any one of various field ordering mechanisms or topological defect models. Generically these give larger CMB anisotropies (from the so-called integrated Sachs-Wolfe effect) for the same density perturbation amplitudes ⁽³⁾. The power spectra of galaxy perturbations, or even the underlying dark matter fluctuations, are notoriously complicated to calculate - nevertheless there seems little evidence that the observed power spectrum can be easily reproduced in these sorts of models. Moreover, there now seems to be some consensus in the view that generic defect models produce at most one (broad) peak in the CMB power spectrum [45] and in a place which tends to give a poor fit to current data.

The status of defects vs. the Universe can be summarized in the following three points. Defect models tend to give

the wrong matter power spectrum;
the wrong CMB power spectrum;
the wrong normalization of matter relative to CMB.

But apart from that, these models seem to work fine!

There is certainly a strong motivation for working on such models simply from the point of view that they are cool. Some of the required mathematics is interesting in its own right and some of the numerical calculations are challenging. It would be neat if the Universe was full of a network of cosmic strings, containing within them trapped regions of GUT-scale physics, and thrashing around at near the speed of light. To put it another way, who wouldn't sometimes rather be Captain Jean-Luc Picard? But ultimately it doesn't matter what sort of Universe we would like to live in ⁽⁴⁾ we are stuck with this particular one, and we are learning a great deal about its properties. Details of the structure within our Universe seem relatively easy to fit with inflationary-type models, and considerably harder with defect-type models.

Which is not to say that defects are not important in other branches of physics - or even perhaps for other purposes in the early Universe - but at this point they seem to hold little promise as a method of forming structure.

³ This is basically because of their similarity to isocurvature models; adiabatic (i.e. inflationary) models, on the other hand, give the correct value to a factor of 2, without really breaking sweat. Back.

⁴ My own sense of humour makes models with say Omega _tot = 1.1 sound pretty appealing! Back.