How to go from the KS theorem to experimentally testable noncontextuality inequalities

Abstract

The purpose of this talk is twofold: one, to acquaint the wider community working mostly on Bell-Kochen-Specker contextuality with recent work on Spekkens’ contextuality that quantitatively demonstrates the sense in which Bell-Kochen-Specker contextuality is subsumed within Spekkens’ approach, and two, to argue that one can test for contextuality without appealing to a notion of sharpness which can needlessly restrict the scope of operational theories that could be considered as candidate explanations of experimental data. Testing contextuality in Spekkens’ approach therefore extends the range of experimental scenarios in which contextuality can be witnessed, and refines what it means to witness contextuality in the presence of inevitable noise in KS-type experiments. We will see this for both KS-uncolourability based logical contradiction type proofs of the KS theorem a la Kochen-Specker and statistical proofs on KS-colourable scenarios a la KCBS or Yu-Oh. While Bell-KS contextuality can be mathematically understood as an instance of the classical marginal problem, the same is not true of Spekkens' contextuality. The latter reduces to the classical marginal problem only under very specific conditions, being more general otherwise. All in all, we will argue that all you really need is Leibniz, i.e. identity of indiscernables, to make sense of contextuality in the most general context.