In discussions of quantum mechanics (QM), a classical system observes a quantum system. There is an asymmetry that observation occurs in only one direction, this is at very different from Newtonian Mechanics when a force applied from A to B implies another force from B to A.

Can one discuss in QM, a system A that is observing B but which at the same time B is observing A?

  • 1
    @conifold: The point I'm making is we always in Classical Mechanics look at reciprocal actions this isn't the case the standard examples of the QM. There may be have been people who have studied this though, I'm just not aware of it. – Mozibur Ullah Feb 15 '18 at 0:59
  • 1
    I am not sure I follow. When we observe planets through a telescope we hardly care about the effect this would have on the planets. The same with measuring temperature or pressure, technically one can not do that without affecting what is measured, but with little care this can almost always be neglected. Bohr's early interpretation of QM ascribed quantum uncertainty to unavoidable observer interference absent in CM, so if anything the asymmetry would go the other way (but this interpretation is now known to be non-viable). – Conifold Feb 15 '18 at 1:23
  • 1
    @conifold: Sure, there the asymmetry is so large that we can neglect the back-reaction this has. Nevertheless Newtonian mechanics is symmetric in the way it talks about force. If we talk about one atom observing another it seems unsuprising we can ask the question does the latter observe the former. Of course QM doesn't talk about such small systems as observers per se, but here I'm taking my cue from relational QM where any system can be considered as an observer. – Mozibur Ullah Feb 15 '18 at 1:30
  • 1
    If decoherence was only applicable to it's own philosophy of QM, why would quantum computer scientists be so worried about their ability to build large scale quantum computers, error correcting algorithms and hardware, etc? That's the kind of decoherence that is relevant to this question. And to the larger point, Conifold is correct that the reason it's always presented as a classical system looking at a QM system is because we are classical systems and the instruments we use are classical because we're macroscopic, not because QM particles can't measure other QM particles. That's decoherence. – Not_Here Feb 15 '18 at 7:45
  • 1
    The many-worlds interpretation solves the asymmetry in your example. > Many-worlds implies that all possible alternate histories and futures are real, each representing an actual "world" (or "universe"). Symmetry is established since "A observes B" has a symmetric partner world where "B observes A" – draks ... Feb 15 '18 at 13:19

The many-worlds interpretation (MWI) solves the asymmetry in your example.

Many-worlds implies that all possible alternate histories and futures are real, each representing an actual "world" (or "universe").

Symmetry is established since "A observes B" has a symmetric partner world where "B observes A"

I admit I favor the "collapse of the wave function" interpretation, but please feel free to pick your own favorite interpretation here.

Back to the question in the title "Why is there an asymmetry in QM?", I'd like to add the following:

Observing a quantum system (made up of a superposition of eigenstates), collapses the wave function to a single eigenstate of your observation operator and by that destroys superposition. More important: the observer A gathers information about the system B*. The (asymmetric) information flow from B to A drives change in the world...

*: This can be shown, because the information can be erased (by so called quantum erasers) while traveling to the observer A. By that, the superposition state is not affected.

  • I'm not a big fan of the many-worlds interpretation but as I've already mentioned I like this answer! It might make me think again about many-worlds... – Mozibur Ullah Feb 16 '18 at 8:48

Especially during the initial formation and development of QM, as far as what physicists write down in their notebooks as “measurements”, they are all classical outcomes. So having a theory that allows one to predict classical observations make sense. Several of the developers of the theory were in the thrall of logical positivism, and thus they were only concerned with constructing a theory that was empirically adequate for computing the results of the kinds of experiments that they could conceive of.

More recently study of quantum decoherence and related phenomena are providing a more complete QM description of the types of interactions that yield essentially classical results. My impression is that from some people in that field you hear the problem as being more of "how does one get an effective classical theory given that the 'real world' is quantum mechanical"?


You seem to have got off on the wrong foot, by comparing observing, to force. For an example of a classical account of the effects of observation, look at https://en.m.wikipedia.org/wiki/Maxwell%27s_demon The key is accounting for information transfer and persistence, not force.

The next problem, is the nature of the superposition of states. An observer, by definition, collapses such a superposition into a definite state by taking an observation. That doesn't neccessarily stop the 'observer' being subatomic, or quantum mechanical, just not in a superposition. Quantum superpositions can interact with each other, making a combined superposition. Observation always means collapse of the wave function and information out into the wider world about it's state.

Also, you are using the word symmetry in a suspect way. You can say Newton's third law states a symmetrical reaction. But symmetry in general related to forces, is likely to be taken as gauge symmetry.

  • 1
    I actually think 'observation' is closer to the notion of force than of observation; but here I'm not thinking of Newtonian force but abstract Aristotelian notion of force which is merely that which causes change. I don't see why the observering system can't be in a super-position with respect to the observed system, though admittedly that isn't a situation that is usually discussed in textbooks. – Mozibur Ullah Feb 15 '18 at 11:36
  • Yes you have made that clear. But your unconventional use of these words has mainly led to lots of confused questions. The nature of quantum mechanics is linked to what the small possible probe is, the information carrier that can interfere least while gathering information. A system, a superposition, must be small enough that no information carriers enter or exit, so that it acts like a black box, the state invisible until a probe acts on it, opening the box. The scale of coherence is limited by how large a system can be isolated from all probes. – CriglCragl Feb 15 '18 at 13:29
  • @craiglcragl: Which unconventional use? – Mozibur Ullah Feb 15 '18 at 17:41
  • Force, observation, symmetry. – CriglCragl Feb 15 '18 at 17:57
  • the uses I put those concepts to in the question are conventional: 'a force is applied from A to B', 'a classical system observes a quantum system', that Newtons third law is symmetric is standard. So I'm not sure why you think I'm using these words 'unconventionally'. Sure, I used them unconventionally in the comment above, but I indicated this. – Mozibur Ullah Feb 15 '18 at 19:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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