Who/What counts as an "observer" in Quantum Mechanics?

Rather than hijack this thread as the OP was on about something different I thought I’d start a new thread (been awhile since I posted so I hope this is ok).

In that thread the following was posted:

Like many people I find this explanation difficult to square with intuition. I realize that my everyday macro experiences are not applicable in most of the quantum world but there just the same.

The double-slit experiment described above I cannot help but get hung-up on the notion of the photon interferring with itself. I realize that it is best described as a wave and not a point particle but even then I cannot see how this works. While the photon my have a probability of being in different locations my notion is that, measured or unmeasured, the photon is never in more than one place at once. While I may not know its location with precision it is still in one location. If we opt for an idea that the photon is, in actuality, in ALL locations simultaneously then don’t you run into problems of the energy of the system? I mean the photon magically multiplies itself millions of times to be everywhere allowable and once measured those duplicates magically go poof?

Further, wouldn’t the photon actually follow a well determined path and not some random “I might be here, I might be there” route? Granted as observers we cannot know everything with enough precision to say with certainty what path the photon will follow but if we magically could know everything about the system in which the photon is travelling couldn’t you then say with precision precisely what route the photon would take? I mean in the end it still should abide by physical laws that govern what happens as it runs into an atom here or a double-slit there no?

And on to this magical observer that collapses quantum wave functions. Who/what is this thing with all this power? As the photon travelling across the room presumably it will run into some atoms or dust particles or the scientist studying this. Does collapsing a wave function require consciousness from the observer? Or put another way, why doesn’t the atom the photon hit not count as an “observer” that can say, “Hey! I know the photon is here since it just hit me!”?

And last but certainly not least is Schroedinger’s poor cat. Why is the assumption the cat is in some undefined state of dead and alive till the researcher opens the box to check? Doesn’t the cat know and count as an observer?

I brought up some of the same stuff here, nearly three years ago. Generated some interesting discussion.

[thread=262540]A Schrodinger’s cat question[/thread]

Just love to mention that, since it was one of the best threads I ever started I think. :smiley:

Friends of Wigner

The “Observer” is a guy with a white face and dark circles under the eyes and dressed in a hooded robe who carries a gelatine “brain” in an oversized Petri dish. everyone knows that!

Actually, lots of mystical types make much about the need for an intelligent Observer to collapse such wavefunctions. It’s part of the Circle of Life!

In fact, QM types think that anything that will result in a macroscopic physical effect counts as an “observer”. I think that “Schroedinger’s Cat” invokes too much hyperbole. Even if the cat weren’t alive – if you used a photograph to record whether or not your vial of prussic acid broke – the permanent effect would constitute an Observer. There are plenty of cases of physical effects in nature that are the result of a wve state being collapsed to yield an unambiguous effect.

In fact, wouldn’t the particles of the cat interacting with each other and with the particles of the box result in a physical cat being present at all times?

Just to muddy the waters further…

Doing a bit of reading it does sound as if the photon is indeed in multiple locations at once otherwise how else could it interfere with itself in a double-slit experiment? Still finding that one difficult to grasp.

Imagine an experiment where we have a photon source and two detectors setup such that the detectors are a light year apart and a light year from the photon source. Now release a photon. A year later is there a chance BOTH detectors will register a hit from that photon? You only released one photon but now you are recording two. If one detector registers the photon first does the wave state collapse literally instantaneously across the whole universe so the other detector cannot register a “hit”? Neat little FTL phenomena if so but I suspect if so the next thing I’d be told is that this method of FTL can carry no information and is useless and not in vioation of anything.

The whole two-slit experiment is already pretty muddy. In some two-slit experiments they reduce the intensity to so low a level that they’re certain that they only get single photons flying through, yet over time the characteristic pattern of interference emerges. So you don’t need for a second photon to be present to create interference effects – the photon “interferes with itself”.
heck, you don’t even need double slits – you can see interference effects from a single slit with finite width – the far-field pattern has the characteristic “sinc function” shape, even when fluence is so low that only single photons go through at a time.

Viewed in that light (HA!) you can see that the insistence on two photons/two slits is causing an un-needed confusion. You could as easily do a three-slit experiment (a triple slit has a characteristic interference pattern that is different from the double slit. There’s a graph in A.P. French’s book “Vibrations and Waves”. Other books have it, too) — would you think that it’s necessary to have THREE photons interfering to see the characteristic pattern? No, you can see it with two (even though that leaves one slit that no photons obviously pass through), or with one. That two photons/two slits thing has needlessly confused generations of physics students. As I say, you can see interference effects with a single slit.

And just to make things more interesting, it’s not clear that Thomas Young ever actually performed that double-slit experiment. I’ve read his original papers, and he never said that he actually did it – it might simply have been a gedankenexperiment. Furthermore, his description of the experiment is somewhat different than the way the experiment is usually presented in books and lectures. But those descriptions do capture his intent and the point he was making.

This is known as the EPR Paradox. You’re in good company in thinking that it’s fundamentally weird (the ‘E’, after all, stands for Einstein.)

I must be missing something with that paradox. For all the talk of waveforms collapsing, and probabilities, I don’t see anything truly paradoxical in it: a pair of electrons is generated, which have opposite spin. When observer A measures the spin of one to be, say, positive, of course the other one will be negative - they have opposite spin. All the observation is doing is showing which particle is which.

I presume the part I am missing is this:

But I am still not really sure why this is considered “spooky”. Even if we can’t know the spin on a given axis, surely we already know that the two particles have opposite spin, on both axes? All we are doing when observing is discovering which is which :confused:

Edit: OK, having read the Wiki talk page, I think I see it… if the x spin is measured first, then the z entanglement is destroyed, and subsequently if particle A is observed to have spin z+, then particle B will be observed to have spin z- or spin z+ with 50/50 probability. Is that it?

I may be in over my head here, but I think the spookiness is because you are assuming a common-sense interpretation: That the particles have a defined spin at the moment they are created.

That seems reasonable, but strict adherence to QM says that this is not the case, only that the spins must be opposite when observed. In fact, things like the double-slit experiment with single photons make physicists question this underlying assumption.

This was the point of Einstein’s paradox: Some interpretation of QM data had led physicists to question basic assumptions about observation in science. Specifically, they note that nowhere in QM is it required that a physical value be determined or determinable before it is observed. That gets them out of some of the clear paradoxes of known experiments (such as the aforementioned double-slit), but as Einstein showed with his thought-experiment, it means that you must assume that once one end of his twin-particle spin experiment is determined, the other end must be instantly determined as well (QM says categorically that the two particles must have opposite spin), and that means this “information” (i.e. the “collapsing” wave function) had to travel at faster than the speed of light.

One way out of the connundrum is to say that, although the wave function collapse can occur at FTL, this posited FTL motion can’t carry information (i.e. you can’t use it to communicate at FTL speeds). So it’s essentially a useless FTL process, having no impact on useful physical phenomenon, so QM can safely ignore it (since QM is only interested in observable effects). It’s not a great solution (and someone around here must know much much more about it than I), but there it is.

Please, I’m a learner like everyone else around here; if my interpretation is incorrect, please let me know (like I need to ask around here :slight_smile: )

Sure, but the idea of the Schrödinger’s cat experiment (which is a Gedankenexperiment or thought experiment) is that it exists a priori as an isolated system, and therefore still in a state of coherent superposition of discrete states (eigenstates) until observed by an external observer. Even though the system is a vast collection of individual states all interacting with each other, the system is predicated on the strictly probabilistic (and easily calculated) state of an individual atom, and thus, the entire state exists simultaneously as two (or equivilently two) recognizable and discernable states.

Now, this obviously isn’t a realistic experiment you could perform for many reasons. For one, it’s entirely impossible to isolate the system inside the box from that outside the box; if you hear the cat meowing or scratching on the box, then you have in effect “observed” the state of the system and thus impacted the result. (Similarly, any attempt to use non-classical methods to identify the state of the system, such as telepathy or clarivoyance, fail to avoid the same issue; it makes no difference whether you open the box or uses some mystical power, you’ve still observed the system and waveform collapse occurs.)

This of course, begs the question the o.p. is asking; to wit, who gets to be the observer and why is he or she so privlidged? Erwin was speaking in the abstract about observing the box–presumably any interaction with the outside world counts as observation and results in decoherence on the larger scale–but many people have, mistakenly in my opinion, taken this to mean that a conscious observer has to perceive and cognate the state of the system before waveform collapse can be said to observed. Again, this begs the question, and Eugene Wigner asked this question explicitly with his Wigner’s Friend Gedankenexperiment, in which Wigner steps outside the room while is friend checks on the cat. Now, Wigner argues, the cat, the box, his friend, and the room all exist in a similarly coherent state of superposition: live cat/happy friend or dead cat/sad friend. (Clearly the friend is more of a lover of cats than physicists of the era.) From Wigner’s point of view, nothing in the universe exists in a “real” state until observed by consciousness, later dubbed the consciousness causes collapse interpretation.

This, in turn, begs another question, or rather two: what the hell is this consciousness thing, and why does it have an effect on waveform collapse? (We’ll neatly dodge the issue of what “waveform collapse” actually is by saying both that we don’t know and it doesn’t matter, which is a cheat that ignores the elephant, but whatever.) The fact is that we don’t have much of a handle on what consciousness actually is–neuroscience being even today in the rather primitive state of trying to identify what areas of the brain are most active when you smell a lavender bush and somehow relate that to the conceptualization of lavander–and thus any definitive connections you make between one area of almost complete unknowns and another is ascientific and about as useless as a bag full of nutrinos, which hasn’t stopped philosophers, New Age gurus, and somewhat confused pop-(pseudo)science authors (including some reputable physicists) from making a lot of money drawing unsubstantiated conclusions between quantum mechanics, religious beliefs, “dancing Wu Li Masters”, and similar nonfalsifiable phenomena.

In fact, there’s no evidence that quantum behavior has any effect on consciousness beyond being the probabilistic basis for chemical reactions, and consciousness itself, insofar as we can say anything about it at all, isn’t a single discrete process but a vast collection of different cognative processes operating on different levels to create the resultant glurge that we think of as “being aware”. Even in the cat gedanken, while the cat won’t be aware that it is dead, it will most certainly have some notion of being alive, and thus the system isn’t seperate from “observation”. (We can assume by default that Wigner’s friend is conscious of his own state.) Therefore, the waveform collapse in the “conscousness causes collapse” interpretation is relative to the conscious observer, meaning it’s either relative to the observer, or the observer is the cognative center of the universe. Either way the result is solipsistic, and as Bertrand Russell would note, this is a fundamentally useless philosophy from which to assert anything.

About that whole waveform collapse event that we’ve studiously avoided discussing: nobody, including all the people in Denmark who bandied it around so freely, have any notion to exactly what this waveform really is and why it collapses like Southern belle when observed. The whole thing is stated as a mathematical formulism which, by definition, can’t be observed in process, merely inferred by result. If this strikes you as the same kind of voodoo that got O.J. Simpson an acquital, then you’re not missing anything, and while it is the most common interpretation taught in basic classes in modern physics (mostly because it directs students to stop worrying about the meaningless and unfalsifiable interpretations and just do the damned math, which works regardless of what you speculate is happening), from what I’ve read alternate interpretations that don’t involve fuzzy concepts about collapsing waveforms are becoming more popular among scientists in the field.

Stranger

What you’re proposing is a hidden variable interpretation, which naïvely has a lot of appeal to it. It takes the Bell Inequalities to show why it doesn’t work, which are unfortunately sufficiently difficult to explain that I won’t even try. The short version is that you can predict statistically the number of each possible outcome that you should get, and in some situations, the numbers predicted by any hidden variable theories disagree with those predicted by standard quantum mechanics. Experiment, unfortunately, agrees with the weird quantum mechanical results, rather than the conceptually-simpler hidden variables, thereby robbing us of a nice, neat understanding of the world.

I didn’t know that’s what I was doing :slight_smile:
I appreciate that I was missing something; after all, if it stumped Einstein there must have been something in it!

Too Late!
Dumb down disproof of “Hidden Variables” for me

Right and what I was trying to get across is that anything that interacts with a quantum entity is an observer. Things can’t by observed without some interations between the thing and the observer.

Two atoms of oxygen in a molecule are each “observing” the other and so the molecule exists someplace real and having physical attributes.

But Bell’s inequality (as I understand it, which isn’t as well as I’d like) only applies to local realism. If you allow for nonlocal hidden variables (and are prepared to accept that entangled particles or widely seperated areas of a waveform potential are connected instantaneously), then it works fine, and in fact Bell was a cautious enthusiast for Bohm-type interpretations which permit a deterministic, objective mechanic with nonlocal realism. This doesn’t verify or falsify this interpretation, but it isn’t invalidated by Bell’s inequality.

He also had an interesting take on the Schrödinger’s cat problem in which the two states of the cat are being hungry or not hungry, rather than being dead or alive. Og bless the Irish.

Stranger

Here’s my version of the OP: Does a Geiger counter have consciousness?

In other words: Let’s start with an assumption that a Geiger counter does not have consciousness. It is also my (possibly incorrect) assumption that when a Geiger counter clicks, it has observed the decay of an atom, which is a quantum event.

If the decay is indeed a quantum event, then it has not definitively occurred unless it is observed by a consciousness. Therefore, if there is no consciousness around listening to the Geiger counter, it has neither clicked, nor has it not clicked. If a tape recorder (which also has no consciousness) is set up to record those clicks, then the recording will be neither blank, nor will it be not-blank, until a consciousness sits down to listen to it.

On the other hand, if someone wants to insist that the tape recorder did have some sort of definitive recording even before a consciousness came to listen to it, then the Geiger counter and/or tape recorder must count as legitimate “observers”.

Can someone please find the flaw in all this?

Mmmm…not according to the Copenhagen interpretation, or at least a broad reading of it. From the perspective of one oxygen, the other could be considered collapsed, but from someone or something that hasn’t yet observed the diatomic system, the whole molecule is just one more complex waveform of two superimposed oxygen atoms. If you could create a situation in which the condition of one atom was predicated on the state of another in a statistical way (analgous to Schrödinger’s Cat) then you’d still have the superposition of states arranged exactly per that statistical relationship. Since we can’t observe waveform collapse–or assume that it has happened before we interact with the system–the entire system, and indeed any reality outside what we’re immediately interacting with “doesn’t exist” except as a collection of possibilities. Hence, no local realism, i.e. stuff isn’t there when you’re not looking at it. (I’m using the term “looking at it” not literally but to represent more generally observing/interacting with it.) Furthermore, “the observer” or the tools used for measurement are assumed to be classical in their treatment, rather than also being systems of quantum particles.

Where does the boundary come between what is decohered into a single, collapsed waveform and what still exists in a formless state of probability? That all depends upon which flavor of interpretation you prefer, and since you can’t look at it before you look at it (or examine states that are nonlocal) there is no effective way to compare the different arguments of what it means to “observe” or “collapse” a waveform, leading to the exclaimation, “Shut up and calculate!” Bohr himself said, “There is no quantum world. There is only an abstract physical description. It is wrong to think that the task of physics is to find out how nature is. Physics concerns what we can say about nature,” indicating that he believed the interpretation to be a mathematical formalism, not a literal event.

I’d like to believe that all of this waveform collapse business is pure bunk, and that there is some more grounded and preferably deterministic mechanic underlying the whole system. But that either requires nonlocal causality–ostensibly a problem for relativity in concept, though in practice the problem goes away–or some other way ontological explanation which resolves the apparently probabilistic behavior with decoherence and entanglement. There’s no particular reason to believe that one interpretation is more likely than the other, though.

Stranger

Even if we assume that consciousness isn’t involved, the box, the counter, the tape recorder, and the tape still exist as quantum systems (claims of classical recording instrumentation be damned) and the whole thing is one big smear of probability until you play back the tape, or at least retrieve it from the room. When collapse occurs is still a big fat question mark.

If we assume that “consciousness” (whatever that is) is somehow actively involved in the process (setting aside the negligable links between consciousness and quantum mechanics–essentially, that we know very little about either of them) it argues an essential solipsistic argument. Such an argument can’t be disproven, but is useless in any material sense. When the world exists at your behest (or the behest of your “unconscious mind”) then there’s no reason to believe that there are any inviolate, fundamental principles or mechanics.

Either way, there’s no way to inspect the system without collapsing it, and we can’t say what state it was in before it collapsed. As a practical matter, we gain no insight into which interpretation is “more right”.

Stranger

My apologies; I didn’t mean to imply otherwise. One can always construct a nonlocal hidden variable theory consistent with any set of observations.

As for the question of when the wavefunction collapses, one could also make the case that it never collapses, and that when I observe the cat, I also exist in a superposition of states, 1/sqrt(2) mourning and 1/sqrt(2) not-mourning. Think about this too much, and it starts looking a lot like the many-worlds interpretation. Which, in turn, should start to provide some insight into why physicists say that the choice of interpretation doesn’t matter.