2
$\begingroup$

The standard explanation for the observer effect is that a small amount of energy from the observer causes the wave function of lets say an electron to break down and for it to develop determinate characteristics rather than probabilities. However why doesn't this happen anyway baring in mind that until its observed it has a probability field spanning the entire universe. It is therefore literally in contact with that energy the observer imparts in the act of observation anyway along with everything else, and yet that doesn't magically cause the probability field to break down. Why only when there is someone who can know things at the other end of this magic energy does an electron take on specific characteristics that (by massive coincidence) can be known?

$\endgroup$

2 Answers 2

5
$\begingroup$

The concept you're looking for is decoherence. Indeed, it happens whether or not an observer is looking. For quantum computer builders this is unfortunate: it would be so much easier if all they had to do to prevent wave function collapse was to avoid looking.

The connection with observation is that observers are incoherent physical systems: decoherence is a necessary feature of the coupling of the observer to the observed. But many other incoherent physical systems exist, and interaction with any of them will produce decoherence. Coherent quantum behavior is only seen in systems isolated from such interactions.

$\endgroup$
5
  • $\begingroup$ Veritasium has a good explanation of decoherence in the context of Schrodinger's Cat. It is part of his explanation of the "Many Worlds" interpretation of quantum mechanics. Parallel Worlds Probably Exist. Here’s Why $\endgroup$
    – mmesser314
    Mar 11 at 13:36
  • 2
    $\begingroup$ @mmesser314 On the basis of many experiments and observations, we know that decoherence is a reasonable model for what we see in reality. But whether parallel worlds exist is a vacuous question scientifically: there is no experiment that can test this hypothesis. $\endgroup$
    – John Doty
    Mar 11 at 13:52
  • $\begingroup$ True. I was trying to say that it is a good explanation of decoherence wrapped in other things that are not part of what is being asked. Never the less, it is a good explanation of decoherence. $\endgroup$
    – mmesser314
    Mar 11 at 13:55
  • $\begingroup$ @mmesser314 It is not true. Decoherence is never complete, so a small remnant could (in theory) always be measured. Besides there are many other examples of things that we cannot see where we still don't think that their existence is a "vacuous" question. This is simply an overly polite attitude towards people who refuse to believe QM. (OK, if your thesis advisor is one of those, it could be wise of course...) $\endgroup$ Mar 11 at 14:31
  • 2
    $\begingroup$ It is objectively ridiculous to posit an infinitely splitting multiverse to answer why quantum measurement is weird. Whether such a notion appeals to you for some reason or not, it is not possible to settle empirically and thus it is not a scientific question. Worse yet, it might be the most abuse that Occam’s Razor has ever gotten. $\endgroup$ Mar 11 at 15:15
2
$\begingroup$

The standard explanation for the observer effect is that a small amount of energy from the observer causes the wave function of lets say an electron to break down and for it to develop determinate characteristics rather than probabilities.

There isn't a standard explanation. Interpretations that involve wavefunction collapse say nothing about the mechanism by which the collapse happens, or propagates through spacetime. It's simply asserted, as an axiom. Interpretations that don't involve wavefunction collapse, like the 'Many Worlds' or Everett Interpretation, do offer an explanation, but neither the interpretation nor the explanation are universally accepted by physicists.

For a brief, hand-waving explanation of how the Everett Interpretation deals with the question, see here. It's not specifically a contribution of energy that triggers the split, but interaction. The observer can provide energy to the system, or absorb energy from the system, or interact in some way that leaves the energy the same. Briefly, if the system observed is in a superposition of orthogonal states, and the observer interacts with it, then after sufficient interaction time the observer shifts to a similar superposition of orthogonal states. 'Orthogonal' means the states do not interact, and thus cannot perceive one another.

The step to classical probabilities hasn't been fully worked out yet - the following is another hand-waving explanation of roughly how we think the argument will go. When the observer system is big and complex, and has zillions of possible states, the split is into a multitude of orthogonal states. In some fraction of them, one outcome is observed, in the rest, the other outcome is seen. By Pythagoras, the squared length of a multidimensional vector is the sum of squares of the individual amplitudes, so if we assume that the total magnitude is conserved, and the individual components are of roughly equal magnitude, then the fraction of the zillions of observer states seeing each outcome will conform to the squared amplitudes of each component of the system's superposition, and will behave like classical probabilities. This is called "decoherence".

However why doesn't this happen anyway baring in mind that until its observed it has a probability field spanning the entire universe. It is therefore literally in contact with that energy the observer imparts in the act of observation anyway along with everything else, and yet that doesn't magically cause the probability field to break down.

For collapse-based interpretations, the wavefunction does collapse, but it collapses to a state virtually identical to the original. The difference between a wavefunction that extends to infinity with an exponentially small probability around the observer, and a wavefunction extending to infinity that is zero close to the observer is imperceptible. The probability field breaks down to zero near the observer, and jumps up slightly by an imperceptible amount everywhere else. It makes no difference to the experimental predictions, so we commonly ignore it, or pretend it doesn't happen.

It can make a difference if the probability is high enough. Imagine a radioactive atom surrounded by detection screens at various ranges. The atom decays and pings off an electron whose wavefunction makes a spherical wave. It hits the closest screen, and immediately collapses to either a detection - where the rest of the spherical wave abruptly vanishes - or a non-detection - where the wave vanishes over the closest screen and abruptly strengthens everywhere else. As the sperical wave passes each detector, it collapses again and again, until the electron is detected. It has to happen this way to guarantee that there will be exactly one detection. The electron cannot be observed at two places at once, and it cannot disappear entirely. And if the detector screens are far enough apart, there is no time for the 'collapse signal' from one detector to reach any of the others in time to tell them what to do.

That is to say, a non-detection (with no exchange of energy) can collapse the wavefunction just as much as a detection does.

In no-collapse interpretations, the split into the superposition of mutually-invisible orthogonal states happens, but one of the components is much smaller than the others. It looks the same to observers.

Why only when there is someone who can know things at the other end of this magic energy does an electron take on specific characteristics that (by massive coincidence) can be known?

The idea that conscious awareness of the outcome is needed to trigger wavefunction collapse was (I think) part of the early development of quantum mechanics, but I think is mostly rejected nowadays.

Collapse-based interpretations are pretty vague about exactly what triggers a collapse. Decoherence is widely thought to play some role - but how you can get non-linear behaviour from still-linear equations is mysterious. Or there are explicitly non-linear theories of collapse, which are in theory testable, but which currently are pure speculation with no evidence for them.

The Everett Interpretation holds that when an observer interacts with a system in superposition, that particular observer enters the same sort of superposition, but other observers don't. So if two electrons interact, then that constitutes observation of one electron by the other, and the electron sees only one outcome, but we as an outside observer see both electrons in entangled superposition.

The appearance of "collapse" - seeing only one possible outcome, picked at random - only happens when you are in a superposition of orthogonal states. Only observers perceive the collapse. To everyone else, the system and its observers are still in superposition, and they will only perceive the collapse when they make an observation on the observers.

It's essentially the same question as asking whether a single electron passing through one slit of the double-slit experiment "sees itself" simultaneously passing through the other. If they did, they would repel electrostatically, and that would change the interference pattern. Since there is no change to the interference pattern, the electron passing through one slit does not "see" the alternatives, and thus as far as each component of the electron's wavefunction is concerned there is only one outcome and the wavefunction has already collapsed. However, this is an illusion caused by being inside the superposition. The component orthogonal states are mutually invisible.

The Everett Interpretation works by taking the linear, local, realist, deterministic, reversible evolution of the wavefunction that everybody accepts applies at the microscopic level in the absence of observation, and points out that if we ask what it would look like from the inside, we predict the appearance of classical physics. So there is no need to posit wavefunction collapse as an axiom. The observations it is meant to explain are already explained. However, since the alternative components of the superposition are forever unobservable, we have no way of proving they don't magically disappear. Thus, this is an 'interpretation' and metaphysics. In the absence of evidence, do you prefer zillions of forever-unobservable alternative 'worlds', or mathematically ugly and forever-unobservable mechanisms to make them disappear? There is no experiment we can do to distinguish them. Which you prefer is down to aesthetics and utility.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.