CosmoCoffee

This paper essentially emphasises the (well-known) fact that single-field slow-roll inflation results do not hold in general if there is more than one scalar field, and discusses a particular setup. They argue that this makes gravity waves more likely to be detectable because T/S after inflation can be larger than you'd estimate for a single field model. However it's not clear to me what T/S are actually plausible? (e.g. if natural two field models give exponentially small tensor ratios anyway then their argument wouldn't really help)

The paper repeats an argument that having \Delta\phi ~ M_p (needed for easily detectable gravity waves in single-field simple cases) is difficult for renormalizable particle theories. However I thought (e.g. from Linde hep-th/0402051) loop corrections depended on the effective mass coming from the potential V(\phi), and hence only required V << M_p^4 rather than a constraint on \Delta\phi?

Note: the quoted requirement for detectable CMB B-modes of r > 10^{-3} may be unduly pessimistic: more optimal delensing can in principle do orders of magnitude better as shown in astro-ph/0310163.

Antony Lewis wrote:
The paper repeats an arugment that having \Delta\phi ~ M_p (needed for easily detectable gravity waves in single-field simple cases) is difficult for renormalizable particle theories. However I thought (e.g. from Linde hep-th/0402051) loop corrections depended on the effective mass coming from the potential V(\phi), and hence only required V << M_p^4 rather than a constraint on \Delta\phi?

I believe there is still something of a split in opinion about whether the requirement is [tex]V \ll M_p^4[/tex] or [tex]\phi \ll M_p[/tex]. So I suppose the argument here is just that [tex]\Delta\phi[/tex] is rolling over a Planck distance and therefore under the more restrictive choice the effective field theory describing inflation won't be valid over the whole range, because before we get to [tex]\phi \sim M_p[/tex] we ought to have integrated in new degrees of freedom belonging to new physics.

I think the issue of loop corrections isn't settled; there is a claim (gr-qc/9609026) that loop corrections lead to an effective screening of the cosmological constant and therefore that such effects are important (astro-ph/0505236) but perhaps the authors didn't intend to invoke loop effects explicitly.

In the "N-flation" paper out last week (hep-th/0507205), the authors give a specific string-theoretic example of this objection (Sec. 2.1, p. 4) based on the idea that as the field gets a Planck distance away from any given minimum of the potential, string scale modes can become light. These light degrees of freedom control the low-energy physics, so the effective field theory changes.

Unfortunately they don't give a reference for this idea. Does anyone have more details? I suppose the scalar is supposed to be interpreted as an inter-brane distance and as one pushes it away from the potential minimum, the branes separate, so at large distances there is a tower of winding modes which are becoming light.

The new paper astro-ph/0509015 gives a nice fairly detailed critique of this paper. They argue that getting any damping of the adiabatic modes is very hard, and give some additional discussion of the \Delta\psi issue.