10.1 Entangled path markers and coincidence sorting
10.1.1 Tagging interferometer arms with orthogonal marker states
Suppose we entangle the two paths \(\lvert L\rangle\) and \(\lvert R\rangle\) of an interferometer with two orthogonal states \(\lvert H\rangle\) and \(\lvert V\rangle\) of an ancillary marker system (such as the polarization of the photon itself or of a twin herald photon). Because \(\langle V\vert H\rangle=0\), the marker carries complete which-path information, and the unconditional signal counts at the interferometer output become completely flat (\(P(0)=1/2\) for all \(\phi\)), showing zero interference visibility.
10.1.2 Sorting signal arrivals by diagonal marker outcomes
Now, instead of measuring the marker in the path-distinguishing basis \(\{\lvert H\rangle,\lvert V\rangle\}\), rotate the marker analyzer to the diagonal superposition basis \(\lvert\pm\rangle=(\lvert H\rangle\pm\lvert V\rangle)/\sqrt{2}\) that projects equally onto both path labels. When each signal detection is paired in coincidence with the marker outcome \(+\) or \(-\), the two sorted sub-ensembles recover full-contrast interference fringes—shifted in phase by \(\pi\) relative to one another!
10.1.3 Complementary sub-ensembles versus the flat marginal sum
In the interactive plot below, three curves appear simultaneously: the blue fringe \(P(0\mid +)\) conditioned on marker outcome \(+\), the dashed gold antiphase fringe \(P(0\mid -)\) conditioned on marker outcome \(-\), and the flat dotted violet line \(P(0)=1/2\) representing all unsorted signal detections. Quantum erasure never alters the flat unsorted signal distribution; it extracts complementary conditional correlations by cross-tabulating signal and marker records.
Visual animation · Section 10.1
Entangled path markers and coincidence sorting
Interactive visual model for 10.1 Entangled path markers and coincidence sorting (Tagging interferometer arms with orthogonal marker states · Sorting signal arrivals by diagonal marker outcomes · Complementary sub-ensembles versus the flat marginal sum). Use Animate or Step to follow the physical state evolution and quantitative predictions.
10.2 Basis rotation of the marker and antiphase fringes
10.2.1 Rewriting the path–marker state in the diagonal basis
Before the recombining beam splitter, the maximally entangled path–marker state is
Substituting \(\lvert H\rangle=(\lvert +\rangle+\lvert -\rangle)/\sqrt{2}\) and \(\lvert V\rangle=(\lvert +\rangle-\lvert -\rangle)/\sqrt{2}\) rewrites the exact same state vector in the diagonal marker basis without any physical disturbance to the signal arm:
10.2.2 Opposite-phase conditional fringes and exact marginal cancellation
After the signal modes \(\lvert L\rangle\) and \(\lvert R\rangle\) recombine at the output beam splitter \(H\), the conditional probabilities of a signal click at port \(D_0\) given marker outcomes \(+\) and \(-\) are
independent of \(\phi\). The crest of the \(+\) fringe sits directly over the trough of the \(-\) fringe, so their sum is identically flat.
10.2.3 Commuting local measurements and post-hoc coincidence sorting
When the signal and marker live on spatially separated photons, their local measurement operators \(M_{\mathrm{sig}}\otimes I\) and \(I\otimes M_{\mathrm{mark}}\) commute: \((M_{\mathrm{sig}}\otimes I)(I\otimes M_{\mathrm{mark}})=M_{\mathrm{sig}}\otimes M_{\mathrm{mark}}\). Consequently, the joint probability \(P(0,\pm)\) is completely invariant under whether the marker is measured before or years after the signal photon hits its detector. Delayed-choice quantum erasure simply sorts already-recorded signal timestamps using the classical log of later marker outcomes.
Visual animation · Section 10.2
Basis rotation of the marker and antiphase fringes
Interactive visual model for 10.2 Basis rotation of the marker and antiphase fringes (Rewriting the path–marker state in the diagonal basis · Opposite-phase conditional fringes and exact marginal cancellation · Commuting local measurements and post-hoc coincidence sorting). Use Animate or Step to follow the physical state evolution and quantitative predictions.
10.3 No-signalling and coincidence-window discipline
10.3.1 Why quantum erasure cannot transmit a faster-than-light signal
Can an experimenter at the marker station send an instantaneous message to the signal station by choosing whether to measure the marker in the \(\{\lvert H\rangle,\lvert V\rangle\}\) basis or the \(\{\lvert +\rangle,\lvert -\rangle\}\) basis? No: in both cases, the signal observer who lacks access to the classical marker outcomes observes the exact same flat marginal probability \(P(0)=1/2\). Extracting either interference fringe requires receiving the classical bit stream of marker outcomes \(\pm\) over an ordinary communication channel.
10.3.2 Reporting both conditional subsets and accidental coincidences
A rigorous quantum-eraser report displays both conditional curves \(P(0\mid +)\) and \(P(0\mid -)\), their relative weights \(P(+)\) and \(P(-)\), the flat unsorted sum \(P(0)\), and the coincidence time window used to pair signal and marker clicks.
10.3.3 Distinguishing coherent basis rotation from deleting classical files
Quantum erasure requires projecting the marker onto the coherent superposition states \(\lvert\pm\rangle\) before the marker decoheres into a macroscopic environment. Merely erasing a computer file after measuring the marker in the \(\{\lvert H\rangle,\lvert V\rangle\}\) basis leaves the signal in an incoherent mixture and never restores fringes.
Visual animation · Section 10.3
No-signalling and coincidence-window discipline
Interactive visual model for 10.3 No-signalling and coincidence-window discipline (Why quantum erasure cannot transmit a faster-than-light signal · Reporting both conditional subsets and accidental coincidences · Distinguishing coherent basis rotation from deleting classical files). Use Animate or Step to follow the physical state evolution and quantitative predictions.
Explore the experiment
Switch among the sorted \(+\) subset, the sorted \(-\) subset, and the unsorted marginal, and sweep \(\phi\) across \([0^\circ,360^\circ]\). Verify that the blue (\(+\)) and dashed gold (\(-\)) curves oscillate \(180^\circ\) out of phase and average to the flat line \(1/2\).
Interactive model · Idealized predictions
Sort the same signal records
On a small screen, swipe the diagram to inspect its components. Moving marks illustrate the model; they are not measured trajectories.
Read the values as a table
Where are the fringes hiding when path information is available? The text and worked examples remain available without JavaScript. Animation starts only when requested; Step gives a motion-free way to follow the model.
A worked example
At phase \(\phi=0^\circ\), calculate the joint probabilities \(P(0,+)\) and \(P(0,-)\) as well as the unconditional probability \(P(0)\) at detector \(D_0\).
At \(\phi=0^\circ\), the conditional probabilities are \(P(0\mid +)=(1+\cos 0^\circ)/2=1\) and \(P(0\mid -)=(1-\cos 0^\circ)/2=0\). Multiplying by the marker probabilities \(P(+)=P(-)=1/2\) gives the joint probabilities \(P(0,+)=P(+)P(0\mid +)=1/2\) and \(P(0,-)=P(-)P(0\mid -)=0\). Summing over the two marker outcomes gives the unconditional marginal \(P(0)=P(0,+)+P(0,-)=1/2+0=1/2\).
Summary
When path-System entanglement marks which arm a particle took, the unconditional marginal signal distribution is a flat sum of two complementary fringe patterns \(\cos^2(\phi/2)\) and \(\sin^2(\phi/2)\) that add to \(1\).
Measuring the marker in the erased diagonal basis \(|\pm\rangle\) and sorting signal detections by coincidence outcome recovers full-visibility fringes (shifted by \(\pi\) between \(|+\rangle\) and \(|-\rangle\)).
Because the total marginal signal remains completely flat without classical coincidence sorting, quantum erasure never transmits faster-than-light messages.
Try it yourself
Problem 10.1
Why can’t the signal observer detect whether the distant marker experimenter chose the \(\{\lvert H\rangle,\lvert V\rangle\}\) basis or the \(\{\lvert +\rangle,\lvert -\rangle\}\) basis without receiving a classical message?
Hint
Sum \(P(0,+)+P(0,-)\) and compare it with \(P(0,H)+P(0,V)\).
Worked solution
In both marker bases, the opposite-phase interference terms \(\pm\frac{1}{4}\cos\phi\) cancel when summed over the marker outcomes, leaving the exact same marginal probability \(P(0)=1/2\) at the signal detector.
Problem 10.2
Suppose a detector measures the marker in the path basis \(\{\lvert H\rangle,\lvert V\rangle\}\) and writes the result to a hard drive, which is later destroyed without being read. Does destroying the hard drive restore interference fringes?
Hint
Ask whether destroying the classical log projects the marker onto \(\lvert\pm\rangle=(\lvert H\rangle\pm\lvert V\rangle)/\sqrt{2}\).
Worked solution
No. Once the marker has been measured in \(\{\lvert H\rangle,\lvert V\rangle\}\) (entangling with macroscopic apparatus degrees of freedom), discarding the classical file simply averages over \(H\) and \(V\), leaving the flat marginal \(P(0)=1/2\).
Problem 10.3
Why must a quantum-eraser publication present both conditional fringes \(P(0\mid +)\) and \(P(0\mid -)\) alongside the unsorted sum rather than plotting only \(P(0\mid +)\)?
Hint
Consider how the two complementary fringes explain the flat unsorted marginal.
Worked solution
Plotting both antiphase subsets \(P(0\mid +)\) and \(P(0\mid -)\) together with their subset rates and unsorted sum verifies that the fringes arise from complementary coincidence sorting rather than from residual single-photon path coherence or biased postselection.
What to carry forward
When path alternatives are entangled with a marker, unconditional fringes vanish while complementary conditional fringes \(P(0\mid\pm)=(1\pm\cos\phi)/2\) remain encoded in joint correlations that can be extracted only by coincidence sorting with a diagonal-basis marker record.
Sources and further exploration
These readings provide historical context, experimental implementations, or deeper derivations. The interactive display above is an educational model, not a reproduction of their measured data.