A proposed inflation falsifier and its loophole

Question: Could a large squeezed/local primordial non-Gaussian signal decisively rule out inflation?

Primary source: Xingang Chen, Hassan Firouzjahi, Mohammad Hossein Namjoo and Misao Sasaki, “A Single Field Inflation Model with Large Local Non-Gaussianity”, Europhysics Letters 102 (2013), 59001. Its introduction states the once-common claim that large local non-Gaussianity would exclude all single-field inflation; the paper then explicitly constructs a non-attractor single-field example with large local signal. The standard consistency relation needs additional assumptions (including constancy of a superhorizon perturbation, i.e. an attractor phase). Thus even a detection would challenge an important class, not by itself all single-field models, much less all inflation.

Where observation stands in the checked collaboration report: Planck 2018, IX: Constraints on primordial non-Gaussianity reports f_NL^local = −0.9 ± 5.1 (68% statistical uncertainty), consistent with no signal. BICEP/Keck’s 2024 programme overview reports the BK18 limit r_0.05 < 0.036 (95% C.L.) and projected sensitivity σ(r) ≲ 0.003 using data through the 2027 observing season; that sensitivity is a forecast, not a measured result. A non-detection of tensors bounds inflation’s scale or disfavours high-r variants, not inflation generally. Do not call BK18 the most recent of all independent combinations: BK18 + Planck PR4 + BAO found r < 0.032 with different data/assumptions.

Use: A particularly crisp distinction between ‘testable model’ and ‘falsifiable whole family’. Paired with Dru’s precise question and the Steinhardt–Guth dispute.