Delayed Choice Phenomena in the Projection Evolution Model

This paper proposes that delayed-choice experiments in a Mach-Zehnder interferometer can be explained within the projection evolution model by treating time as a quantum observable, where the phenomenon arises from the temporal overlap between the photon's wave function and the interferometer devices.

The Gist

The Big Idea: Time is a Place, Not Just a Clock

In standard physics, we usually think of time as a rigid clock ticking in the background. A particle moves through space, and the clock just tells us when it happens. But the authors of this paper argue that this view is incomplete. They propose a model where time is treated just like space.

Imagine a movie reel. In the standard view, the film plays frame by frame, and we watch it happen. In this new model, the entire movie reel (past, present, and future frames) exists as a single, solid block. A particle isn't just a dot moving across the screen; it is a "blob" that stretches across both the screen (space) and the length of the film reel (time).

Because the particle has a "length" in time, it can overlap with events that happen before or after its "center" arrives. This is the key to solving a famous quantum puzzle.

The Puzzle: The Delayed-Choice Experiment

To understand what the paper solves, imagine a classic game called the Mach-Zehnder Interferometer. Think of it as a fork in the road for a photon (a particle of light).

1. The Setup: A photon hits a splitter (like a traffic light) that sends it down two paths at once.
2. The Choice: At the end of the paths, there is a second splitter.
   • If the second splitter is present, the two paths recombine, and the photon acts like a wave (interfering with itself).
   • If the second splitter is removed, the photon acts like a particle (taking one specific path).

The Mystery: Scientists have shown that you can decide to put the second splitter in or take it out after the photon has already passed the first splitter but before it hits the detectors.

• The Paradox: How does the photon "know" whether to act like a wave or a particle if the decision is made after it has already started its journey? It seems like the future is changing the past.

The Paper's Solution: The "Ghostly" Overlap

The authors say: "No time travel is needed." Instead, they look at the temporal profile of the photon.

Imagine the photon isn't a tiny, hard marble. Instead, imagine it's a long, fuzzy cloud or a sausage that stretches out in time.

• The "head" of the sausage might be at the first splitter.
• The "tail" of the sausage might be way back in the past or way forward in the future.

The Analogy of the Foggy Hallway:
Imagine you are walking down a hallway (the photon) that is covered in thick fog (the time profile).

• If someone puts a wall in the hallway (the splitter), and your foggy tail is still touching the spot where the wall will appear, you will bump into it.
• It doesn't matter if your "head" (the main part of you) hasn't reached the wall yet. Because your "fog" is already there, the wall affects you.

The paper claims that in the Delayed-Choice experiment, the photon's "time-fog" overlaps with the moment the scientist decides to insert or remove the splitter.

• If the splitter is there while the photon's time-fog is overlapping it, the photon behaves like a wave.
• If the splitter is gone during that overlap, the photon behaves like a particle.

The photon doesn't need to "know" the future. It just interacts with the setup based on how much of its "time-body" is touching the equipment at that moment.

The Three Scenarios Tested

The authors ran computer simulations (mathematical models) with three different shapes of "time-fog" to see how the photon would react:

1. The Box (Symmetric): Imagine the photon is a perfect, sharp-edged box of time. It interacts with anything that overlaps its edges. If the splitter appears while the box is passing, the interaction happens.
2. The Tail (Asymmetric): Imagine the photon is a comet with a long tail.
   • If the tail points backward, the photon "feels" changes made in the past before its main body arrives.
   • If the tail points forward, the photon "feels" changes made in the future after its main body has passed.
   • This explains why a decision made after the photon passes the first splitter can still change the outcome: the photon's "tail" is still hanging around the second splitter when the decision is made.
3. The Gaussian (Realistic): This is a smooth, bell-curve shape (like a normal distribution). It shows that even with a smooth shape, the overlap between the photon's time and the device's time determines the result.

The Conclusion

The paper concludes that we don't need to believe in "retrocausality" (the idea that the future changes the past). We just need to accept that time is a dimension the particle occupies, not just a clock we watch.

• Old View: The particle is a point; time is a line. The future can't touch the past.
• New View: The particle is a "spacetime sausage." It stretches across time. If the setup changes while the sausage is overlapping it, the sausage reacts.

By treating time as a quantum observable (something you can measure and interact with, just like position), the "delayed choice" mystery disappears. It's simply a matter of temporal overlap, just like a long train overlapping a station platform.

Technical Summary

1. Problem Statement

The paper addresses fundamental interpretational issues in non-relativistic quantum mechanics regarding the nature of time.

• The Core Issue: In the standard Schrödinger formulation, time ($t$) is treated as a classical, external parameter rather than a quantum observable. Consequently, the wave function is "quantized" in space but not in time. This asymmetry creates difficulties in formulating a consistent space-time description of quantum processes.
• The Pauli Obstacle: The Pauli theorem generally forbids the representation of time as a self-adjoint operator in the standard Hilbert space $L^2(\mathbb{R}^3)$, preventing a direct operator-based treatment of time evolution.
• The Delayed-Choice Paradox: Wheeler's delayed-choice experiment demonstrates that a photon's behavior (wave-like vs. particle-like) seems to depend on the experimental setup (presence of a beamsplitter) chosen after the photon has entered the apparatus. Standard explanations often rely on the spatial width of the wave function reaching the point of change. However, the authors argue that a consistent explanation requires treating time as a quantum observable to account for temporal overlap between the photon and the experimental devices.

2. Methodology: The Projection Evolution (PEv) Model

The authors utilize the Projection Evolution (PEv) formalism to reformulate quantum dynamics.

• Four-Dimensional Formulation: Time ($x^0$) is treated as a component of a four-position vector $x = (x^0, \vec{x})$, on equal footing with spatial coordinates. Time is represented as a self-adjoint operator $\hat{x}^\mu$ with eigenvalues corresponding to time coordinates.
• Evolution Parameter ($\tau$): Since time is an observable, it cannot serve as the evolution parameter. Instead, evolution is described by a discrete index $\tau$ (evolution steps). The state evolves via mappings $E|: \mathcal{K}_\tau \to \mathcal{K}_{\tau+1}$.
• State Definition: The wave function $\psi(x)$ depends on all four spacetime coordinates. The evolution is governed by:
  $$\psi(\tau + 1; t, x) = \frac{E|_{\tau+1}\psi(\tau; t, x)}{\|E|_{\tau+1}\psi(\tau; t, x)\|}$$
  where $E|$ can be unitary (Schrödinger-like) or projection operators.
• Temporal Uncertainty: The model introduces a temporal uncertainty $\Delta t$ paired with a temporal momentum uncertainty $\Delta p_0$. Localization in time occurs via interaction (projection), similar to spatial localization.

3. Application: Mach-Zehnder Interferometer

The authors apply the PEv model to a photon traversing a Mach-Zehnder interferometer with time-dependent beamsplitters.

• State Space: The system is modeled in a two-dimensional spacetime ($t, x$) with two orthogonal channels ($|1\rangle, |2\rangle$). The state is $\bar{\Psi}(t, x) = \psi(t, x) \otimes \sum c_k |k\rangle$.
• Photon Wave Function: The photon is described by a separable wave function $\gamma(\tau; t, x) = \gamma_t(\tau; t)\gamma_x(\tau; x)$.
  • Spatial Part: Modeled as a shifted spherical Bessel function (approximated as a sinc-like function) representing semiclassical motion.
  • Temporal Part: Defined via Fourier transform of a frequency profile $a(\omega_t)$. The authors test various profiles:
    • Symmetric: Gaussian or rectangular (box) functions.
    • Asymmetric: Functions with temporal tails directed forward or backward in time.
• Interaction Mechanism: Beamsplitters ($BS_1, BS_2$) are modeled as projection operators $P(\Omega_\ell)$ that act only when the photon's spacetime coordinates overlap with the spacetime region $\Omega_\ell$ where the device is present.
  • If the photon's temporal profile overlaps with the device's presence time, the state is mixed (interference occurs).
  • If there is no temporal overlap, the state remains in its original channel (no interference).

4. Key Results

The authors calculate detection probabilities for detectors $D_1$ and $D_2$ under various delayed-choice scenarios:

• Symmetric Case (Gaussian/Box Profile):

  • The photon interacts with the setup based on the spacetime overlap of its wave function and the device.
  • If the device is present during any part of the photon's temporal profile, the interference pattern is established.
  • The results match classical expectations: maximum detection probability occurs at the spacetime location of the detector.

• Asymmetric Case (Temporal Tails):

  • Backward Time Tail: A photon with a tail extending into the past reacts to changes made before the photon's main peak reaches the device. It "knows" about the setup configuration prior to its arrival.
  • Forward Time Tail: A photon with a tail extending into the future reacts to changes made after its main peak has passed the device. This allows the photon to be influenced by a delayed choice made after it has seemingly passed the interaction point.
  • Scenario 2 (Forward Tail): When $BS_2$ is inserted after the photon's peak passes the spatial location of $BS_2$ but overlaps with its forward temporal tail, the second detector ($D_2$) registers a small probability of detection, indicating interference was induced by the late insertion.

• Scenario 3 (Dynamic Insertion):

  • When devices are inserted and removed dynamically, the detection probabilities shift based on the precise temporal overlap between the photon's profile and the device's existence window.

5. Significance and Conclusions

• Resolution of Delayed Choice: The paper demonstrates that delayed-choice phenomena do not require retrocausality (influence from the future to the past). Instead, they are explained naturally by the temporal overlap of the photon's wave function with the experimental apparatus.
• Redefinition of Causality: In the PEv model, causality is not a strict point-to-point relationship in time. Between localization events, the object is "smeared" over a time interval. Causality exists only between points where the object is localized on the time axis.
• Time as an Observable: By treating time as a quantum observable, the model provides a consistent framework where the "decision" of the photon to behave as a wave or particle is determined by the total spacetime interaction history, not just the spatial configuration at a specific instant.
• Implications: This approach unifies the description of temporal and spatial quantum processes, offering a potential solution to the "time problem" in quantum mechanics and providing a new perspective on quantum clock construction and time-of-arrival measurements.

In summary, the authors successfully model the delayed-choice experiment by shifting the focus from spatial wave function width to temporal wave function width, showing that the photon's interaction with the setup is governed by the four-dimensional spacetime overlap, thereby resolving the apparent paradox without invoking non-local retrocausality.