At some point Dirac writes:
(OpA)(OpB)Y = 0
where OpA and OpB are those two brackets that differ only in the sign of m, then he deduces:
(OpA)Y = 0 OR (OpB)Y = 0
(or is that AND).
I don't get this. If he'd started with:
(OpA)Y . (OpB)Y = 0
then fair enough, but that's not the same as (OpA)(OpB)Y = 0
(OpA)(OpB) is just Klein-Gordon, right, so that can't be what he means.
Perhaps he means that he had the choice of writing:
(OpB)(OpA)Y = 0
and that the whole thing wouldn't be zero unless the result of applying the right hand operator alone was zero, but why should that be?
Why can't the right hand operator applied to Y output something non-zero that just so happens to go to zero under the left hand operator, and vice versa?