4.4.4. Power-law spectra

It is currently fashionable to assume that the perturbation spectra, especially the scalars, take on a power-law form, just as it was fashionable pre-COBE to assume the Harrison-Zel'dovich form. Both these cases are approximations to the true situation, and as observations improve the power-law approximation may be found wanting just as the Harrison-Zel'dovich one was post-COBE. Staying within the single-field paradigm, unless prominent features are dialed into the inflationary potential, deviations from power-law behaviour are however predicted to be small, perhaps even by the standards to be imposed by the Planck satellite. Nevertheless, one should test for deviations when improved data come in [23, 24, 25].