(3) The KS theorem, by its mathematical nature, is not empirically testable. However, we could, along the lines of the previous paragraphs, try to measure a subset of a suitable KS-uncolourable set. Especially, it should be possible to produce cases along the lines of Clifton’s example (3.5) where QM and a noncontextual HV theory make measurably different predictions. It seems as if such cases could provide empirical tests of whether Nature is contextual (though not whether such contextuality is