A Conservative Theory of Semiclassical Gravity
We argue that semiclassical gravity can be made consistent if quantum systems source gravity only when they participate in non-gravitational interactions that lead to environment-induced decoherence. Outside such decoherence-based events, systems do not contribute their stress-energy to the semiclassical equations, so regions lacking these interactions may remain (approximately) flat. The proposal is testable by probing the gravitational field sourced by systems, which should depend entirely on environment-induced decoherence; by gravity not mediating entanglement in the Bose-Marletto-Vedral (BMV) experiment; and by how the reversibility of the initial state in this experiment would depend solely on this decoherence, distinguishing it from competing approaches. We propose a kind of decoherence-inducing interaction that leads systems to source gravity: it models decoherence as chains of causally ordered, localized interactions between quantum matter fields, selecting the states and observables that source gravity. We argue that these interactions lead to the emergence of gravity. One way to see this is to note that these chains consist of timelike and lightlike separated events, whose causal order determines the metric up to a local conformal factor (Hawking-King-McCarthy-Malament theorem), and that when the number of events can be associated with the four-volume of spacetime, it provides the remaining information to fix the metric. Another way to understand this emergence is to note that these events can be correlated, so that the metric can be understood as arising from these interactions. This framework is conservative: it does not modify standard quantum theory while providing a consistent semiclassical theory of gravity. It may also explain why the vacuum need not gravitate, predict a time-varying ``cosmological constant,'' and provide a semiclassical estimate of its value.