Someone famous in the field of philosophy of mind (although I’m not at liberty to say) asked me the following question regarding my most recent blog post on the logical contradiction of quantum mechanical reversibility:
If one can't prove that Schrodinger’s Cat was in a superposition, I presume the same goes for “Schrodinger’s Particle.” But we seem to get that evidence all the time in interference experiments. Are particles different in principle from cats, or what else is going on?
What we do in a typical double-slit interference experiment is start with a bunch of "identically prepared" particles and then measure them on the other side of the slits. The distribution we get is consistent with the particles having been in a linear superposition at the slits, where the amplitudes are complex numbers. The fact that they are complex numbers allows for "negative" probabilities, which is at the heart of (the mathematics of) QM.
The key is that no particular particle is (or can be) observed in superposition... rather, it's from the measurement of lots of identically prepared particles that we infer an earlier superposition state.
The problem is that it's technologically (and I would argue, in-principle) impossible to create multiple "identically prepared" cats. If you could, you would just do lots of trials of an interference experiment until you could statistically infer a SC state. But since you can't, you have to rely on doing a single experiment on a cat, by controlling all its degrees of freedom, so as to reverse any correlations between the cat and the quantum event. But by doing so (assuming it was even possible), there remains no evidence that the cat was ever in a SC superposition at all. So, since science depends on evidence, it's not logically possible to scientifically show that a SC ever existed... and no one seems to have addressed this in the literature.
Amazingly, this paper just came out in Physical Review Letters, so it's something that people in the physics community are just now starting to wrap their heads around. The paper doesn’t go far enough, but it at least points out that if WF makes a “measurement” but then is manipulated to show that WF was in a superposition, then even that “measurement” has no results.