A Schrodinger's cat question

I’ve been listening to the audible.com narration of ‘dirk gently’s holistic detective agency’, and there’s a bit on the Schrodinger’s cat thought experiment that has started me wondering, so I’m putting my question up here for any quantum physics gurus to weigh in on.

First, a recap of the S C formulation described in the book. You have a large lead box (presumably with a large enough air supply to keep a cat going for a few hours under ordinary circumstances,) that can be sealed to entirely enclose the space within. You have a cat. You have a lump of some very weakly radioactive material that wouldn’t harm the cat directly. You have some sort of geiger counter rigged up to a vial of poison so that if the counter detects a radioactive particle, the poison will spread throughout the air in the box in sufficient concentration to kill the cat.

You seal the box up for a few hours, over which time you’ve calculated that the odds of a radioactive emission that would set off the geiger counter are 50%. That’s the setup.

Now, the classic line of reasoning goes as so forth… because of the uncertainty principle, the position of the subatomic particles involved in the chunk of radioactive material are indeterminate until they are observed. Specifically, a particle may have two different waveforms, one in which it remains stable and one in which it assumes the characteristics of a radioactive emission. Since many of these partial-waveform particles will strike the counter, you end up with the remarkable situation in which the entire cat develops two waveform patterns, superimposed on each other, one in which the cat is alive and one in which it is dead. Once you open the box and peek in, observing the situation you collapse the waveforms and only one of them becomes ‘real’.

(Formulations of the standard spiel will vary - if I’ve made any serious goof up to this point let me know, of course.)

Now, here are my own thoughts. For this to make sense, you have to stipulate that the geiger counter is not itself ‘observing’ the particles… that it’s just one link in a chain reaction that is incapable of collapsing the waveform of the individual particles, but that instead the geiger counter will develop its own duplicate waveforms, if only for a slight instant, as a result of the partial-waveform particles striking it. In the same way, the poison vial does not ‘perceive’ the electric signal from the geiger counter and collapse its waveform, but resonates with alternate waveforms of its own. And the cat, who I think would be capabel of observing the poison gas arriving or not arriving and collapsing that dual waveform into a reality, does not - it develops its own dual waveform itself.

So now, my question is this: if all of the above is true, what makes me capable of collapsing the waveforms by opening up the box and peeking inside? Don’t I also develop a dual waveform, one in which I say “oh look, the cat’s alive” and one in which I say, “Oh look, the cat’s dead.” And presumably the audio recorder running in the lab also develops a dual waveform based on whatever I said and the clipboard develops a dual waveform in which ‘alive’ might be checked off and ‘dead’ might be checked off… and so on.

I presume there isn’t something unique about human beings that we can collapse waveforms but no other creature or device in the universe is capable of doing so… so what’s the deal? Of course, if you believe in the theory that infinite multiple universes are splitting away from each other based on each and every moment, then there is really no ‘collapsing’ of waveforms in the way that the schrodinger’s cat explanations suggest – each alternate waveform simply gets a universe of its own to be real in. But I’d like to hear any possible explanation that doesn’t involve alternate universes.

Thank you in advance for your time and expertise. :smiley:

I’m really just giving you a little bump here, but I’ll weigh in with some thoughts and gladly be humiliated by a physicist later.

As I understand it, the uncertainty principal limits what you can know about what is, it does not limit what is.

In other words, the cat is either alive or dead, but you cannot know which until you observe. The cat is not both alive and dead, it is just that from outside the box it may be as if this were true. There is something going on in the box and the cat is definitely alive or dead, you just won’t know which case is true until you observe.

Next stop, philosophy 201: subjective reality vs. objective reality.

It’s a gedanken experiment illustration regarding the limits of observation at the intersection of nature and quantum reality, not a Heathkit “Feline Transmogrifier”. Trying to move this experiment out of the thought experiment stage essentially “collapses” everthing.

IANA Physicist, but IIRC the idea of the experiment is that the entire box, including cat, RA material, vial and counter, are a closed system and thus may exhibit a “communal” partial wave functions until the system is “opened” to observation from the outside. No, there is nothing unique about a human being - there could be a video camera inside connected to a time-lapse circuit, so that no human could observe the setup until some time later. In this case, the original system could be said to be a “dead cat-in-a-box” system from the moment of the recording of the death. The expanded closed system, consisting of the whole gizmo described above + the video recorder (closed until someone or something scans the recording) can be said to go into the “system-containing-a-dead-cat” state only at the moment the death of the cat is actually scanned by an external agent (human or non-human)… ad infinitum.
Note that there in no “cat” here. There is a closed system, consisting of the cat + assorted hardware. This system can be said to be in an unknown state - possible states being “system including live cat” and “system including dead cat” - until its state is “collapsed” by external observation.

(Or something like that… I think. If I’m not mistaken. And I probably am.)

As astro said, this is a though experiment about the limitations of physical models, not about radioactivity and cats…

[hijack]BTW, a question to real physicists - is it true that Scroedinger chose a cat for this thought experiment with the punch-line “curiosity killed the cat” in mind?[/hijack]


Cecil’s poetic ramblings.

If you ask me, if you ever open the box, the condition of a “completely sealed box” is violated. Once you open the box, you can determine not only that the cat is dead now, but also that it has been dead for some time, and how long that time is (if you want to nitpick about the precision of forensic methods, then just replace the cat with a very precise clock). So even though the box is sealed as of T = 0.5 hours, you’re still observing what the state of the system is at that time. To really perform the “experiment”, you’d have to seal the box, wait an hour, and then chuck the whole thing unopened into a black hole. Which means that you don’t get any results at all, which makes this a rather pathological experiment.

Quantum mechanic checking in…

Quantum mechanics, as it is usually presented, postulates two different ways for a system to evolve:

  • a “closed” quantum system evolves in a particular, reversible way (this evolution is called “unitary” evolution); and
  • when measured, a quantum system “collapses” irreversibly into one of a set of states defined by the measurement operator.

The question you ask is a long-standing question in quantum mechanics, and there have been a lot of debates and theories and interpretations put forward to try to answer it. Really there are several questions:

(1) Does a “measurement” really collapse the wavefunction?
(2a) If so, what defines a measurement? How is a measurement different from any other quantum-mechanical operation?
(2b) If not, why does it look like it? Why don’t we ever see superpositions of states like (live-cat)+(dead-cat)?

You might want to look at this paper by Max Tegmark (PDF). Among other things he presents the result of a poll on question (1) he took among physicists at a conference. The poll indicates a fairly even split. In my experience (even less scientific than Tegmark’s poll), most theoretical physicists don’t think a measurement really collapses the wavefunction. After all, it seems like every individual interaction between particles in the procedure we like to call a “measurement” seems to be adequately described as a unitary evolution, so the total process seems like it should be as well.

(Tangent: There’s some dispute over whether any physical process at all is theoretically irreversible. A prime candidate (as Chronos says) is with black holes, which classically are described by a very small amount of information (mass, charge, angular momentum) and which therefore seem like very good information-erasing devices. Not everyone buys this argument, though, and some recent theoretical results hint that black holes actually do store large amounts of information along their event horizons, so maybe they don’t really erase information irreversibly.)

Tegmark gives a very good discussion of the many-worlds interpretation (motto: Not quite as weird as you might have heard!) and a popular proposed answer to (2b). The idea is basically that the environment constantly interacts with things and becomes correlated with them (unless they are very well shielded). Because physics is local, the correlations usually happen in such a way that the position of the system becomes correlated with the environment. This is indistinguishable, to a casual observer (i.e., one who does not examine the environment very carefully for correlations with the system), from the system having “collapsed” as a result of a position measurement. (This idea is often called “decoherence.”) It also explains why we never see objects in position superpositions.

I personally believe in many-worlds (the version described by Tegmark), at least for normal phenomena (I’m more agnostic about black-hole unitarity, though it seems much more plausible now than it did ten years ago) so I’m probably not the best person to try to answer (2a). There have been explanations proposed, ranging from the mystical (“intelligence” or “human observer”) to proposed new physics beyond QM. Quantum gravity is often mentioned, though since QG doesn’t actually exist yet it always sounds a little like hand-waving to me. (The idea as I understand it is that, just as in the decoherence theory, different states of the measurement apparatus interact differently with the quantum gravitational field, then a miracle occurs, and then some QG effect causes a wavefunction collapse.)

So what’s the real answer? Well, there haven’t been any experiments with anything like a definitive answer to (1) (of course, a definitive answer to (1) would require an answer to (2a), a definition of “measurement,” as well). The reason is that these are really really hard to perform. Typically, to measure the relevant sorts of quantum effects on particles, they must be kept carefully isolated from all external influences in some sort of trap in the middle of a vacuum chamber. This is hard enough to do for a handful of particles; to do it for any large-scale measuring device is probably a lot harder.

Quantum mechanic, I like that.

Scene: Dirac’s Garage

<The mechanic sucks in a breath> “You see, your eignenvector’s entangled 'round your black-body. I can order in some parts from the manufacturer, and pop those in, but there’s no way of knowing whether it will work or not till we observe it. So it’s impossible to predict the price with certainty, but it’s not going to be cheap, buddy.”

What if you open the box but have your eyes closed so you still don’t know whether the cat is alive or dead (i.e. you have not yet observed it)? Will the waveform not collapse until you open (one of) your eyes?

Omphaloskeptic has already addressed most of the main points here better than I could have. I’ll just add that the Schrödinger gedankenexperiment can be extended to the idea of Wigner’s friend: if my friend puts the gedankencat in the gedankenbox for whatever amount of time, then according to me, she remains in a superposition of “having observed a dead cat” and “not having observed a dead cat” until I call her up on the gedankenphone and ask her what the result was.

Of course, then Wigner started going on about how this proves that consciousness is necessary to make an “observation”, and such talk always makes physicists nervous & uneasy, and most of them said, “That’s nice, Eugene” and went back to thinking about other things.

I had always assumed that the irreversibility criterion meant that when the quantum world interacted with any large ensemble of macroscopic bits (whether it is a gieger counter, cat or brain) then a “measurement” occurred. Is this view incorrect or testable?

As I said above, this is one theory out there, and it is in principle testable. The measurement-postulate experiments I know of are mostly testing what’s called a “quantum eraser.” The idea is to perform the operation that correlates the quantum system with the measurement apparatus, and then to undo this operation before observing the measuring device. The result of the experiment will be different, depending on whether a measurement apparatus performs some irreversible operation (“wavefunction collapse”) simply by virtue of being correlated with a quantum system. The goal is to differentiate this case from the case of environment-induced decoherence, in which such macroscopic correlations don’t inevitably cause collapse but merely tend to do so because such systems typically have very strong interactions with the external environment.

So far this has not been done for a “measurement apparatus” consisting of a macroscopic number of particles, simply because the experimental state of the art is not yet able to do that. It will probably first be done using a Bose-Einstein condensate (which may or may not be large enough to perform the wavefunction collapse, depending on exactly what threshold you take to constitute “macroscopic”) or similar system for the “measuring” device, because such systems are relatively simple to keep confined and coherent. Trying it with a human or Geiger counter is well beyond the current state of the art.

I’ve always had problems with this gedankenexperiment in particular, but also with gedankenexperiments in general.

Specifically with respect to Shroedinger’s Cat, why does everyone seem to forget about the great likelihood that the cat is an observer?

In my experience, felines are particularly observant. And it’s his freakin’ life, after all. Even if he does have eight more, they’re still important to him.

And secondly, it take great, tremendous confidence to use a “thought experiment” as evidence to support a hypothesis. I know they’re merely for illustrative purposes, but they are just as easily used to illustrate falsehoods and mistaken models. For instance, I can think up thought experiments to support and illustrate just about every discounted theory… from Lamarckian Acquired Traits to Conscious Design Theory (or whatever they call Creationism this week).

And specifically, in the case of SC, who’s to say that there’s not some other medium that the lead box is not closed to, that allows some sort of observation of the state of the system. It could be audio (I didn’t hear him go >thud< yet!) or neutrino, or most importantly, some phenomena that is yet unknown to the theorist.

In other words, a gedankenexperiment assumes that the gedanker has perfect, omniscient knowledge of the physical universe.

Yes, a gedankenexperiment is intended to ask questions, not answer them, but as is all too common with scientific concepts ported to popular media, the intent and effect became far too easily distorted.

IANAQP but I understood the purpose of the Schrodenger’s cat was to show how QP views the world. Before QP we could only say that the cat is dead or alive. QP allows for a cat that is 50% dead and 50% alive at the same time - or both dead and alive at the same time. (If there is a one percent chance of the particle decaying - the cat is then 99% alive and 1% dead.) IOW two physicists would be arguing that the cat is either dead or alive when the quantum physicist would come up and tell you that you have a half dead cat in the box (If he was an optimist the cat would be half alive.)

I don’t see the big deal or why this has to be done with a decaying atom at all. Couldn’t we produce the same ‘chance’ results buy building a similar box with cat and poison. But instead of a Geiger counter, have a device that can read dice. Then have a little hole in the top of the box that closes the instant you drop in the dice. If the dice roll a 12, then the machine activates the poison. So is this cat now 1/12th dead and 11/12ths alive??? (Or however many possibilities there are. I know there are s but load) After all, we can’t know for sure what the results of the roll were unless we open the box. What is it about atoms that makes everyone go nuts about this cat? Am I missing something? No one would sit there and postulate all this waveform stuff with dice, would they?

Or to simplify it even further: If I have a cup and I shake the dice around and then flip the cup upside down on the table so I can’t see the dice. Is it the case that just because I can’t see the dice, that they are in a state of every result until I lift the cup and see that the result was snake-eyes? Did I collapse some kind of magic waveform? That would be collapsing a whole bunch of waveforms, wouldn’t it??

Basically, why make the system so complicated in the first place? How does the unpredictability of atoms differ from that of other things. And why was this whole thought experiment started with atoms in the first place? Why not cards?

Yes, it is an observer, but it is inside the system. It knows if it’s alive, but for someone outside of the system (IIRC), it’s in both a live cat and dead cat state until the waveform is collapsed.

Similarly, suppose the cat had tricked the experimenter and slipped the poison into his food (obviously a Siamese cat) and that the amount of the poison was such that 50% of all humans of the researcher’s weight would die within the half-life of the radioactive isotope. The cat goes in, and the scientist begins to eat his sandwich. When the time has elapsed, for the cat, then, the researcher is in a state where he’s 50% alive and 50% dead.

And so if there is an observer inside the system, then the eigenstates collapse, and there is no duality… there is just a 50 percent chance that the cat is alive.

If the cat is an observer, there is no quantum mechanical mystery, just a normal everyday statistical process. Like the “how long until the next bus arrives” problem in probability and statistics 101.

As I perceive the SC concept, it was an attempt to demonstrate how quantum uncertainty can be brought out of the subatomic scale and into the macroscopic scale… among other things.

(Like the differences between superpostion models, and the Einstein-Rosen multiple universes model. In the latter model, there are two new universes created every instant - one where the cat has just died as a result of the decay event happening, and another where the cat is still alive, waiting for dinner.)

But outside of that closed system, i.e., the researchers, the box cannot be treated as a box with either a live cat or a dead cat. They must treat it as a system of the superposition of those two states, which is, from my dim memory of sophomore physics, not the same thing mathematically.

of course the problem is that the “outside observer” must form another closed system to an observer outside him, thus the quantum uncertainty propagates to fill the entire universe - hence the many worlds model or other strange theories (e.g conciousness collapses the states).

I have this probably naive sense that the universe is not so strange (it seems uneconomical that the entire universe has to accomodate one atoms decay). Thus the collapse to me should happen at the level where the quantum interacts with the macroscopic. Whether this can be accomodated by present theories (e.g. by a kind of statistical-thermodynamic type irreversability) will be fascinating.