Can gravity account for the emergence of classicality?

A recent debate has ensued over the claim by Pikovski et al. [Nat. Phys. 11, 668 (2015)] that systems with internal degrees of freedom undergo a universal, gravity-induced, type of decoherence that explains their quantum-to-classical transition. This decoherence is supposed to arise from the different gravitational redshifts experienced by such systems when placed in a superposition of two wave packets at different heights in a gravitational field. Here we investigate some aspects of the discussion with the aid of simple examples. In particular, we first resolve an apparent conflict between the reported results and the equivalence principle by noting that the static and free-fall descriptions focus on states associated with different hypersurfaces. Next, we emphasize that predictions regarding the observability of interference become relevant only in the context of concrete experimental settings. As a result, we caution against hasty claims of universal validity. Finally, we dispute the claim that, at least in the scenarios discussed by Pikovski et al., gravitation is responsible for the reported results, and we question the alleged ability of decoherence to explain the quantum-to-classical transition. In consequence, we argue against the extraordinary assertion by Pikovski et al. that gravity can account for the emergence of classicality.

Figure 1

Trajectories of the wave packets in the $K$ frame.

Figure 2

Trajectories of the wave packets in the $K^{\prime }$ frame.

Figure 3

Trajectories of the wave packets and constant-time hypersurfaces of $K$ and $K^{\prime }$. Observers in $K$ compare proper times between segments $(a,b_1)$ and $(a,b_2)$, finding no difference; observers in $K^{\prime }$ compare segments $(a,c_1)$ and $(a,c_2)$, finding a difference.

Figure 4

Free-falling observers compare proper times between segments $(a_1,b_1)$ and $(a_2,b_2)$, finding no difference; static observers compare segments $(a_1,c_1)$ and $(a_2,c_2)$, finding a difference.

Figure 5

The reported effect disappears if one first evolves the system to a hypersurface of constant $t^{\prime }$ and then back into a hypersurface of constant $t$.

Figure 6

Spacetime diagram of the situation where the centers of mass of the two wave packets are at rest in the corresponding inertial frame and where the proper-time difference is only due to the particular form of the final hypersurface.