= Solution
The <Einstein–Podolsky–Rosen paradox> begins with correlations enabling certain predictions about one system from alternative measurements on a separated partner. Its <EPR criterion of reality> promotes certainty without disturbance to a sufficient condition for an element of reality. Together with a locality premise, the alternative position and <momentum> predictions motivate simultaneously real quantities not represented by one standard <quantum state>. The intended conclusion is incompleteness, not a mathematical inconsistency of the predictions.<Can Quantum-Mechanical Description of Physical Reality Be Considered Complete? by Albert Einstein and others>[1935, pp. 777–780, reality criterion and alternative distant predictions]{ref}
<Niels Bohr> challenges the inference from absence of mechanical disturbance to a context-independent assignment of reality. He accepts that there is no mechanical disturbance of the distant system at the relevant stage. His response invokes the experimental conditions defining the possible predictions: arrangements establishing a spatial reference and arrangements enabling a <momentum> balance are complementary. On this view, changing the local measurement changes the whole phenomenon to which a definite property attribution belongs. It does not remotely kick the other particle. His diaphragm discussion illustrates why specifying the apparatus and its reference frame is part of the physical claim.<Can Quantum-Mechanical Description of Physical Reality Be Considered Complete? by Niels Bohr>[1935, pp. 699–700, diaphragm arrangements and the paragraph distinguishing mechanical disturbance from conditions for prediction]{ref}
A <spin singlet state> gives a clean reconstruction. Its state is
$$
|\Psi^-\rangle=\frac{|0_z1_z\rangle-|1_z0_z\rangle}{\sqrt2}
=-\frac{|0_x1_x\rangle-|1_x0_x\rangle}{\sqrt2}.
$$
Here $|0_x\rangle=|+\rangle$ and $|1_x\rangle=|-\rangle$; the minus sign is an irrelevant overall <quantum phase>. Measuring the first <spin> along either axis predicts the other's same-axis outcome with certainty. But the two alternative local <quantum measurements> define different conditional ensembles. Averaging over the unread outcome leaves the distant <reduced density matrix> $I/2$ in both cases. The contrast between conditional prediction and the unchanged distant marginal is <quantum steering> together with <quantum no-signalling>.
This explains how consistent quantum predictions can coexist with the absence of an operational remote disturbance. It also identifies what a Bohrian rejection of the EPR inference amounts to: certainty in alternative contexts does not license combining the corresponding property attributions into one context-independent account. Merely invoking the <Heisenberg uncertainty principle> would not suffice, since EPR specifically argued that locality permits the attribution before any incompatible distant measurements are jointly attempted.
\b[The reply is satisfactory as an articulation of <quantum complementarity>, but not a demonstration that the EPR locality premise and completeness are jointly sustainable.] The critic can distinguish a change in our permissible predictions from a change in the distant physical condition. If the latter really is unchanged, why must the alternative predictions not reveal pre-existing properties? Bohr's contextual standard rejects that demand rather than deriving its impossibility from neutral premises. Later <Bell inequalities> sharpen the issue: <quantum no-signalling> does not imply <Bell local causality>. A satisfactory contemporary account should acknowledge this difference, rather than crediting Bohr with a proof of Bell-local completeness or interpreting his reply as faster-than-light mechanical influence.
Back to article page
