quantum pathology

December 20, 2007 12:55 PM

quantum physics. what am I not getting?

my understanding of quantum physics is that on the really small scale, our traditional laws of physics aren't so cut and dry so we had to develop probability models for how stuff will behave in order to make testable predictions. is that even approaching accurate?

if its not pretend my question is a request for a layman's explanation for quantum physics. If its close enough, read on for my actual question...

isn't that kind of lazy science?

I mean, my understanding is that in traditional physics, nothing is random. coin tosses, dice rolls, all of it can be predicted if you've got enough data. seems counterintuitive that this would change just because you're getting really small.

appearances aside - it seems to me sort of defeatist to say "stuff happens that we don't expect. there is no way we can know how or why, so lets just gather enough data to be able to say how its probably going to happen next time" Like if we never understood how gangrene happened, but instead of searching for the cause, we developed a probabilistic model for how and when it would strike. quantum pathology? that would never fly.

are we even still looking for the forces or effects that cause discrepancies at the quantum scale?

I don't know, maybe this question is silly and I don't understand the science well enough, but it seems like just declaring something unknowable and moving on isn't very scientific at all.

posted by nihlton to science & nature (48 answers total)

You might want to read about Hidden Variables - I think that Wikipedia entry is about what you are asking , no?
posted by vacapinta at 1:00 PM on December 20, 2007 [2 favorites]

The short answer is the Uncertainty principle and the observer effect.
posted by swordfishtrombones at 1:00 PM on December 20, 2007

Coin tosses and dice rolls aren't physics, they're probabilistic tools. The rate of a falling body, or the effects of friction and force (to name a few topics) are governed by the laws of physics. There's no probability about it. Objects on a non-quantum level behave in very predictable ways.

Very very very tiny things don't follow these rules. For example a quantum particle can be observed to be in two places at the same time. This article on the Uncertainty principle is a bit above a layman's introduction to some of the ideas in quantum physics but is a
posted by brain cloud at 1:03 PM on December 20, 2007

On preview: the page that vacapinta links to seems to be a better answer to your question.
posted by swordfishtrombones at 1:03 PM on December 20, 2007

Depending on how badly you want an answer, here are some askMe's looking for books introducing quantum physics: 1, 2.
posted by shothotbot at 1:04 PM on December 20, 2007

doh...hit post instead of preview.

Anyway...swordfishtrombones has linked the wikipedia article. That's a cool concept to wrap your head around. There's also The Particle Adventure which does an amazing job of describing quantum particles and their behavior. You could probably spend days going through it, and it would be worth it if this is an area of interest for you.
posted by brain cloud at 1:05 PM on December 20, 2007 [1 favorite]

on the really small scale, our traditional laws of physics aren't so cut and dry so we had to develop probability models for how stuff will behave in order to make testable predictions.

No, not at all. Traditional physics made testable predictions at all scales; we discovered that those predictions did not match reality at very small scales (the level of a single electron, for instance). We needed a new theory to describe reality, and quantum physics was it.
posted by xil at 1:06 PM on December 20, 2007

but it seems like just declaring something unknowable and moving on isn't very scientific at all.

There are some physicists who try working out a unity another way around - David Bohm has some interesting theories, for instance, and is generally considered legitimate. At this point, actual experiments show some very weird results, though. It's not just probability; it's that matter acts differently from what we'd expect at quantum scales. The double-slit experiment is a famous example.

But quantum theory is certainly not a "done" science. There is a lot of frustration in the field as people fight over the next step, some people saying we're nearly there, others saying we've headed off on pointless tangents... The most recent potential breakthrough I've heard about was that Garrett Lisi E8 theory. We'll see what comes of it.
posted by mdn at 1:07 PM on December 20, 2007

Is it lazy to say that the gravitation between two bodies is proportional to the product of their masses and inversely proportional to the square of the distance between them? Thats just the way the world works.

We might think it is "strange" or "counter intuitive" but thats because we have evolved in, and our entire brain and all of our senses are calibrated to a much larger scale. I imagine that the way we live would be very "strange" to bacteria (if they could make that judgment).

Physics is not being lazy, it is creating models that reflect reality, and then testing against reality to see if those models hold. If thats lazy blame nature.

As far as "declaring things unknowable" perhaps you should read this. Something things just are unknowable.
posted by stilgar at 1:10 PM on December 20, 2007

For those who aren't adherents of the hidden variables theory, the probability in quantum mechanics represents fundamental non-determinism, i.e. things that we cannot predict because it is impossible to predict them, not because we lack the understanding to predict them.

In this context, classical physics can be predicted because the effects of non-deterministic phenomena are much smaller than the deterministic on a large scale.

IANAPhysicist, but I understand that one way to look at quantum mechanics is that many possible universes are collapsed into a single one by the act of observation... an electron really IS probabilistically in many places at once, until you nail it to an exact spot by looking at it.
posted by qxntpqbbbqxl at 1:13 PM on December 20, 2007

all of it can be predicted if you've got enough data. seems counterintuitive that this would change just because you're getting really small.

My layman's opinion is that you've memorized the facts without getting it. Quantum mechanics says, "In the very small, our framework of classical Newtonian physics does not apply." It's not lazy to say things aren't predictable when they can be in two places at once. You're evaluating statements about quantum mechanics from a Newtonian mindset.
posted by yerfatma at 1:21 PM on December 20, 2007

"Aren't perfectly predictable". Not like I get it either.
posted by yerfatma at 1:22 PM on December 20, 2007

fwiw - I haven't memorized *any* of the facts. I was just curious after having listened to folks talking about q. physics. (my physics educations doesn't extend beyond a semester of physics in highschool)
posted by nihlton at 1:28 PM on December 20, 2007

If you trust relatively simple matrix mathematics, then the uncertainty in momentum and position within the Heisenberg QM formulation is unavoidable. Why the weirdness? Because things down there do not behave like things up here. All our physics models are verbal/linguistic/mathematical metaphors and analogies to map the described behaviour of one thing onto another thing. We say that a quantum object behaves sometimes like a "wave", and at other times like a "particle". These are abstract concepts that we have formulated based on our observation of macroscopic objects relatively close to us in space and time. The truth is that a quantum object behaves like a quantum object. We only found this out very recently. What that means for how things really behave up here or far away is still being found out.
posted by meehawl at 1:28 PM on December 20, 2007

Other people have answered your question more specifically than I can, but I'll add that I disagree with your statement that statistically-based theories, observations and laws are "lazy science."

You're acting as if statistical statements are fuzzy statements. They're not. They're precises. If I say, when I toss a coin 1000 times, approximately 50% of the tosses will be heads, that's not fuzzy, it's accurate. Approximately 50% of the tosses WILL be heads.

It doesn't tell you anything about how a specific toss will go, but that's not what I was talking about. I was talking about the trend over 1000 tosses. If I said something like, "If I toss the coin 1000 times, EXACTLY 517 tosses will be heads," that wouldn't be accurate. My "approximately 50%" is accurate.

Statistics != guessing.
posted by grumblebee at 1:29 PM on December 20, 2007 [1 favorite]

Read the Bell's Theorem article at wikipedia, and in fact start with the section labelled "Importance of the Theorem".
posted by Wolfdog at 1:38 PM on December 20, 2007 [1 favorite]

seems counterintuitive that this would change just because you're getting really small

Yes, it does, doesn't it? That really bothered/bothers alot of physicists.

But, isn't it kind of counterintuitive too that our notions of how physics works should be exactly the same at say, the size of a galaxy, as compared to the size of an electron? There's no good reason for it.

In other words, just because it's counterintuitive doesn't make it wrong. In fact, all sorts of phenomena in physics, both classical and quantum, is counterintuitive-- spend some time fooling with a gyroscope and you'll see what I mean. The problem isn't the physics, it's the intuition. In fact, this is one of the reasons students have a hard time with quantum mechanics when they first start studying it - our natural physical intuition is built up in the classical world, and this intuition is frankly crappy when applied to quantum phenomena.


In short, no, it's not lazy. You don't just say "I don't know where the electron is"-- instead you compute a wavefunction that gives you a "probability as to where it will be" at any given point in time. Next time you solve Schroedinger's equation tell me afterwards how lazy it felt.

Probability, in my opinion, isn't even the right word; I guess I'm personally an adherent of the many-worlds interpretation that qxntpqbbbqxl (that's a mouthful) alluded to. What the square of the wavefunction tells you is the probability you'll make a certain measurement; that is not the same statement as "it's the probability the electron is somewhere".


Oh and mdn, I would not say that Bohm is generally accepted; two weeks ago I was at a conference with alot of quantum fundamentals folk, and there was alot of scoffing when he came up. I'm not an expert in that area so I can't say precisely *why*. Also, re: Lisi, yes his theory is interesting, but a) there are some major problems (e.g. the Coleman-Mandula theorem), and b) it doesn't really have much to say at all about quantum fundamentals; rather it's an attempt to unify various forces.


A resource which y'all might find useful are the public lectures at Perimeter Institute, for this subject particularly these two.


(IAAP, but not a quantum fundamentals physicist)
posted by nat at 1:57 PM on December 20, 2007

I think what you are missing here is that a basic principle of quantum physics is that everything has particle nature in that sub-atomic entites have a specific quanta of mass, and electrical charge. But at the same time, sub-atomic entites also have a wave nature which means that they act more like premiable fields than Newtonian objects. So the Heisenberg uncertainty principle isn't just "particles are unpredictable to human beings." It's that there are many areas when we look at molecular and sub-atomic scales where we can't explain what is happening without treating matter as a wave function.

As an example, understanding covalent bonds in organic chemestry is only possible if you consider electrons as smeared-out regions of electrical force. If the electron in that molecular bond really was a tiny point of charge with an electric force that followed the inverse square law, organic chemistry wouldn't work, and just about everything that we've learned about organic chemistry goes down the toilet.

As another example, Bose and Einstein predicted that at very low temperatures the wave nature of ordinary matter will dominate to such an extent that the waves of individual atoms will overlap, producing a new form of matter with unique properties. This has been confirmed with the creation of Bose-Esinstein condensates.
posted by KirkJobSluder at 1:58 PM on December 20, 2007

"my understanding of quantum physics is that on the really small scale, our traditional laws of physics aren't so cut and dry so we had to develop probability models for how stuff will behave in order to make testable predictions. is that even approaching accurate?"

Not really.

Putting QM in it's historical context: there were a bunch of observations (the ultraviolet catastrophe, the photoelectric effect, and others) that could be explained elegantly if light energy, and subsequently the energy states available to the electrons in an atom, was not allowed to take on arbitrary values but was rather quantized into discrete packets. So people began building models involving these so-called quanta, but they didn't have any fundamental explanation as to what was causing the quantization. Until 1926, when Erwin Schrödinger proposed an elegant equation that, when combined with classical methods for describing the energy in a system, described a quantized reality. And it describes observed reality very, very accurately.

Schrödinger's equation makes use of a mathematical construct called a wavefunction that did not have an immediate and obvious physical interpretation. For a long time, the most popular interpretation of this mathematical construct related it to a probability. This is not the only interpretation, and I'm not even sure if it's currently the most fashionable interpretation. This interpretation essentially allows you to manipulate wavefunctions to derive information about the probability of particles that cannot be directly observed having a certain position and momentum. Heisenberg's uncertainty principle puts constraints on the accuracy of this information.
posted by mr_roboto at 1:58 PM on December 20, 2007

On preview:
Ack, phenomena are.

In other news, what meehawl said, basically.
posted by nat at 1:59 PM on December 20, 2007

"stuff happens that we don't expect. there is no way we can know how or why, so lets just gather enough data to be able to say how its probably going to happen next time"

I think others have provided the links to cover what quantum physics is and isn't, but your statement makes me think that you misunderstand how science (and especially physics) works.

In very broad strokes, the scientific approach is to observe something, develop a theory to model the observed behavior, and then test the theory through experiment. Then we observe something new, or something that doesn't fit our previous theories and so we develop a new theory and test it, and so on. Most sciences, but in particular physics, are reductionist, which means that we try to explain everything in terms of the smallest things possible. This gives the appearance of explaining why things are the way they are, but it doesn't really accomplish this: it only explains things in terms of smaller units. Even if we discover that everything is made up of a certain type of little bits, we still don't have any explanation for why those little bits behave the way they do. However, we have simplified things greatly, because we only have to worry about one type of little bit, instead of all kinds of different things.

Maybe an example would help. Newton sat under that legendary apple tree and was bonked on the head and then developed his theory of gravitation to explain that (and many other) observations. He said that apples fall towards the earth because objects with mass are attracted to each other and provided a handy method to calculate this attraction. He didn't find out the ultimate reason why apples fall towards the earth, he just theorized that something called gravity does it. He didn't tell us how or why it happened. Now, later, we developed atomic theory and we reduced Newton's theory about massive objects to a theory about individual atoms, so we can explain the behavior of apples in terms of the behavior of atoms. Later, Einstein claimed that gravity is actually a deformation of space-time, but that just shifts the why question over. Why is space-time deformed? This same problem is still being worked on today. We still don't have a good explanation for gravity.

Before quantum physics, physicists thought they pretty much had it all figured it out and just had to clear up a few more problems. They thought they could explain everything, but it turns out that they were very wrong. Some of those remaining problems (or in other words, observations that they couldn't explain with their theories) led them to quantum physics. So far, our quantum model seems to explain what we observe pretty darn well, but that might not always be the case.

Now, most physicists would disagree strongly with your claim that they have declared quantum physics unknowable and moved on. Trying to explain quantum physics in terms of something more basic (i.e. in terms of smaller bits) is a big part of contemporary physics, but it turns out to be a really hard problem (one of the most popular theories is string theory, but it still doesn't hasn't made any predictions that we can actually test). A big part of the drive to explain quantum physics in more basic terms is that there are a lot of different little bits in our quantum model. Physicists don't like that and so they try to explain things in more basic terms. They certainly haven't given up in any sense.

Finally, your question seems to presuppose that a deterministic model is inherently superior to a probabilistic model. Some physicists might agree with you, but others aren't willing to accept that. We want our model to predict the behavior of objects in the universe. Maybe the universe actually behaves probabilistically, rather than deterministically, in which case our model should be probabilistic.
posted by ssg at 2:02 PM on December 20, 2007 [1 favorite]

Also, I forgot to mention that when quantum physics was originally being developed, a lot of physicists had the same impression you had, i.e. that a probabilistic model isn't real physics and so must be bullshit. However, once the predictive power of the quantum model became apparent, those sorts of criticisms died down.
posted by ssg at 2:14 PM on December 20, 2007

Here's the thing, we might be able to accurately predict everything on the subatomic scale, but the problem is, we can't measure things accurately enough to account for all the variables, and we can't tell accurately enough where things are. Not only that, we can't even THINK of ways to do it properly, so we have to come up with formulas that get us "close enough".
posted by blue_beetle at 3:05 PM on December 20, 2007

ssg: However, once the predictive power of the quantum model became apparent, those sorts of criticisms died down.

That's an important point. Quantum physics is not a kludge for dealing with the problems of observing things on subatomic scales. (Which is why I cringe every time I see or hear the word "observer.") Nor is it modernist magic that results from extreme experiments. Evolutionary processes have created proteins that make use of electron tunneling, so the effects have an objective impact beyond just the investigations of researchers.
posted by KirkJobSluder at 3:11 PM on December 20, 2007

mean, my understanding is that in traditional physics, nothing is random. coin tosses, dice rolls, all of it can be predicted if you've got enough data.

That's not correct, and that's because it is impossible to "get enough data". The combination of Heisenberg uncertainty and chaos theory says that even extremely tiny errors in your data describing the initial condition can result in an incorrect prediction of the outcome if you look far enough out, because errors in the simulation cascade and multiply, and because it is physically impossible to perfectly measure the positions and velocities, even for large objects.
posted by Steven C. Den Beste at 3:21 PM on December 20, 2007

The book that helped me understand what quantum physics is about is The Dancing Wu Li Masters. It's a layman's introduction, so it never really went above my head, and I've had about as much formal education in physics as you, nihlton (i.e. basic high school stuff).
posted by good in a vacuum at 3:37 PM on December 20, 2007

Oh and mdn, I would not say that Bohm is generally accepted; two weeks ago I was at a conference with alot of quantum fundamentals folk, and there was alot of scoffing when he came up.

I know he's not mainstream - that's why I said "generally legitimate" - I just meant he was credentialled & published and that sort of thing. But yes, if it wasn't clear it's not the prevailing view. On the other hand, I've never heard a real explanation of why his theory is rejected. On the wikipedia page it suggests that the choice between the standard interpretation and Bohm means one has to either reject realism or reject locality, and as it turns out, most scientists prefer to reject realism. This seems like a very weird choice to me.

So, yea, it's not the popular POV, but minds as great as Einstein never could see how the standard quantum model could be right. I mean, 20 years ago, it was thought adult brains couldn't change, and now plasticity is all neurologists are talking about, so who knows where the science on this will go. It seems to me that recent books have been more along the lines of "what now" and "where did we go wrong" rather than "that's all set then", but it may just be a blip of anxiety rather than any evidence of a coming shift.
posted by mdn at 3:39 PM on December 20, 2007

SCDB: That's not correct, and that's because it is impossible to "get enough data".

Sure, but the question here is about a the difference between a deterministic model, i.e. Newtonian mechanics, where the same inputs will always produce the same outputs, and a probabilistic model, i.e. the quantum model, where that doesn't hold. Just because you can't get all the data, doesn't mean that the model isn't deterministic.
posted by ssg at 3:50 PM on December 20, 2007

Correct me if I'm wrong, but don't macroscopic objects follow quantum rules as well? IE classical mechanics is a special case of quantum in which quantities are so large that all probabilities approach 0 and 1.
posted by msittig at 8:05 PM on December 20, 2007

It's not that we don't have good enough equipment to make the measurements that we'd need, or that we just haven't figured out how to yet; the important thing is that quantum mechanics prohibits us from ever making those measurements. It's really not just a matter of compensating for some present limitations.
posted by you're a kitty! at 12:16 AM on December 21, 2007

SSG, when it's possible to create predictive formulas in Newtonian/Einsteinian physics at the macroscopic level, then everything seems to be nicely deterministic. But there are a lot of problems for which that can't be done. The classic example of that is the "three body problem" in orbital dynamics. You can set up the equations -- but you can't solve them, except in special cases. There is no known general solution to the "N-body problem" where N>2.

So the only way to predict the future state in a 3-body problem is "brute force", iterative simulations. And the difficulty there is that these are the kinds of systems which in chaos theory are referred to as having "extreme sensitivity to initial conditions". Set up two identical cases and make a very small change to one of them, and then let them run, and eventually their states will diverge so far that they won't resemble one another at all.

Heisenberg uncertainty (even for objects the size of planets and stars) is enough for that, if you look far enough out. It is impossible to perfectly predict the 3-body case infinitely far into the future using brute force, because it is physically impossible for you to be able to measure the initial state accurately enough.
posted by Steven C. Den Beste at 12:48 AM on December 21, 2007 [1 favorite]

I'm not qualfied to talk about this issue, but ever since I read Jaynes's critique^* of the Copenhagen interpretation I have had similar concerns. The money quote on p 1013:

In current quantum theory, probabilities express the ignorance due to our failure to search for the real causes of physical phenomena. This may be unavoidable in practice, but in our present state of knowledge we do not know whether it is unavoidable in principle; the "central dogma" simply asserts this, and draws the conclusion that belief in causes, and searching for them, is philosophically naive. If everybody accepted this, no further advances in understanding of physical law would ever be made; indeed, no such advance has been made since the 1927 Solvay Congress in which this mentality became solidified into physics.^+
+Of course, physicists continued discovering new particles and calculation techniques — just as an astronomer can discover a new planet and a new algorithm to calculate its orbit, without any advance in his basic understanding of celestial mechanics.

I've never had time to really think about his objections carefully, or look for orthodox rebuttals, but I find his objections at least superfically appealing.


^*see p 1011, "But What About Quantum Theory?" It's part of his wonderful book Probability Theory — The Logic Of Science. Well, I think it's wonderful, but I am a Flaming Bayesian, and he was one of the great champions of Bayesian statistics.
posted by Coventry at 6:19 AM on December 21, 2007

SCDB and you're a kitty!: It seems to me that you both are following a "Where's Waldo" metaphor of quantum mechanics which assumes that somewhere in that plum pudding of electromagnetic force that we call an atom, there are discrete objects floating around in a soup of uncertainty. In the "Where's Waldo" metaphor those particles "really" have a discrete position and momentum, but we just can't measure both with a high degree of precision. The Newtonian 3-body problem is certainly a "Where's Waldo" because in it you still have three gravitational objects that are theoretically simplified to point sources. Stars in tidal streams from galaxy collisions are still stars with a roughly spherical shape and a center of gravity.

It's not that quantum mechanics "prohibits us from ever making those measurements," (of energy and position). Electrons of a fixed energy level simply don't have a discrete position to measure. Instead, we have shaped clouds of electrical force in atoms, or extremely short electron waves in an electron microscope. The wavelength of electrons is on the same order of scale as the bonds between atoms, so in many cases we are forced to talk about electrons as waves and clouds.

So there is a big difference that makes the uncertainty of quantum physics different from the uncertainty of the three-body problem. We can point our telescope at Saturn and say with a high level of confidence that Titan was at this position sometime last night when light reflected from its methane clouds started the journey towards Earth. But no matter how hard we look at DNA under a microscope, we can't spot the electron, and furthermore the structure fails to make any sense unless we give up on the desire to treat the electron as a particle with a position. When we consider the electron as a cloud or a wave, then organic chemistry starts to make sense.
posted by KirkJobSluder at 7:03 AM on December 21, 2007 [3 favorites]

Coventry: Well, working through it, it seems that his description is a bit odd. The photoelectric effect is actually fairly explicit about what properties of light result the emission of electrons. Photons must be of sufficient energy to bump electrons into a free energy state. This is further explained with quantum electrodynamics. Page 112 has what I think is the money quote of his misunderstanding:

In classical statistical mechanics, probability distribution represented our ignorance of the true microscopic coordinates ignorance that was avoidable in principle but unavoidable in practice, but which did not prevent us from predicting reproducible phenomens, just because those phenomena are independent of the microscopic details.

In other words, it is the "Where's Waldo" fallacy that assumes that somewhere in the plum pudding there is a particle that can be modeled with classical formalisms rather than a wave that is better modeled by the formalisms of probability.
posted by KirkJobSluder at 7:36 AM on December 21, 2007

I don't think he's pushing a particle-based theory. How do you get that from that quote? It seems to me he's saying that we are wimping out by assuming that there is no deterministic explanation whatsoever.
posted by Coventry at 7:49 AM on December 21, 2007

Coventry: It's there in "true microscopic coordinates" and the metaphor of the coin toss which assumes that there is a Waldo in the picture.
posted by KirkJobSluder at 8:03 AM on December 21, 2007

"True microscopic coordinates" refers to the model used in classical statistical mechanics, not to a hypothetical deterministic model for quantum mechanical phenomena. The coin-toss example addresses the main thesis of the chapter, that the indeterminacy in probabilistic models generally reflects our subjective ignorance of the systems being modeled, rather than some intrinsic property of the systems. It's not saying anything about what a deterministic model of quantum mechanics would look like.
posted by Coventry at 8:14 AM on December 21, 2007

SCDB:

The solvability of N-body problems doesn't have anything to do with the deterministic nature of Newtonian mechanics. It doesn't matter if you can solve a particular problem in practice or not, it only matters if the same initial conditions will always lead to the same results. Newtonian mechanics is, of course, wrong, but that's how the Newtonian model works.

Your point that the Heisenberg uncertainty in the position of large objects is enough of a perturbation to change calculated results is well taken. As msittig points out above, there isn't actually such a thing as Newtonian mechanics; there is just quantum mechanics on scales large enough that probabilities collapse to near unity. I glossed over that in what I've written above and I probably shouldn't have.

However, Heisenberg uncertainty is part of the quantum model and doesn't appear in Newtonian mechanics, so if you are making an argument by appealing to Heisenberg uncertainty, you aren't actually talking about Newtonian mechanics.
posted by ssg at 8:22 AM on December 21, 2007

Coventry: We very well may be ignorant of the systems we are trying to model with quantum mechanics, but it is still the best model we have so far. We have all kinds of experiments that test the quantum model and they tend to agree with the model.

I don't accept the argument that the universe must be deterministic, because I've never heard it stated in any logical way. It seems to always be just a feeling that someone has. It confuses me when people insist that the universe couldn't be probabilistic. Why? What evidence do they have?
posted by ssg at 8:33 AM on December 21, 2007

The Copenhagen interpretation could well be the truth about what's going on with quantum indeterminacy. However, I understood the OP to be asking whether it's sloppy to assume that it is, and stop looking for a more detailed, deterministic explanation. I was pointing out a physicist who would have agreed with him on this point.

You are correct that there is no evidence that physical laws aren't fundamentally indeterminate, but Jaynes's point was that there is no evidence that they are, either. The fact that we have a probabilistic model for a system may merely reflect our ignorance of the system's fine details.
posted by Coventry at 8:55 AM on December 21, 2007

I'd say it is pretty bad science to look for a "more detailed, deterministic explanation." Why set out looking for a deterministic explanation? Wouldn't it be better to set out looking for the best explanation?

There certainly is some pretty good evidence that the universe is probabilistic: our best model is probabilistic. Yes, that doesn't guarantee anything, but it seems like quite a stretch to insist that we remain agnostic on this point because we might be ignorant of finer details. We might be ignorant about the finer details of a lot of things, but we do the best we can.

My point is that all the arguments for a deterministic universe or for remaining willfully agnostic seem to take the same form as Einstein's "God doesn't play dice with the universe" argument, i.e. the I-don't-like-it-so-it-ain't-true argument.
posted by ssg at 10:52 AM on December 21, 2007 [1 favorite]

Well, I would say it's the Copenhagen interpretation which is willfully agnostic, since it posits a model which is intrinsically indeterminate, and says that there is no hope of ever resolving its indeterminacies. It says that there are some experiments where you don't know why one event occurred rather than another. And refusing to consider the possibility that there is something else to know about why it happened seems pretty willful.

It's conceivable that there is intrinsic indeterminacy in the way the universe evolves. But seeking to explain events which current models can't explain is hardly bad science. You could say Jaynes's position is that it would be better science to accept the agnosticism of contemporary quantum mechanical models only reluctantly, and be open to the notion of searching for more detailed models which can more reliably and precisely predict the outcome of experiments for which current models give relatively uninformative probabilistic predictions.
posted by Coventry at 11:23 AM on December 21, 2007

I don't think you can take a normative approach to determinism, i.e. determinate = good, indeterminate = bad, and be scientific at the same time. We build models that follow from our observations, not our preconceived notions about how the universe works. Cf. KirkJobSluder's "Where's Waldo" argument above.
posted by ssg at 11:35 AM on December 21, 2007

Yes, but we build scientific models to predict and explain phenomena. A model which did a better job of that than contemporary quantum mechanical models would be a better model, from a scientific standpoint, so to turn away from the possibility that such a model is possible when there is no compelling reason to do so is bad science, and unfortunately, that's what the Copenhagen interpretation is doing. The determinism is a red herring.
posted by Coventry at 11:43 AM on December 21, 2007

The assertion that the Copenhagen interpretation somehow prevents the development of any future physics models is ridiculous. Anyone can develop their own model and try to show it can be better experimentally confirmed than the standard model. The continued popularity of string theory over the past 20 years, despite it not having yielded a single testable prediction, is surely evidence of the openness of physicists to other models.
posted by ssg at 12:55 PM on December 21, 2007

Oh, come now. Jaynes obviously isn't saying it prevents the development of future physics models. In fact, he explicitly addresses that in the footnote I quoted. He's saying that the widespread acceptance of the Copenhagen interpretation has led to a dearth of research into more fine-grained models which would sharpen the indeterminate predictions made by the current model. Take Schroedinger's cat experiment, for instance: I don't know much about string theory, but to the best of my knowledge, it offers no better predictions about the time at which a particle will decay. As far as I know, we still model that with a poisson process, and the Copenhagen interpretation says the poisson process is intrinsic to the system.
posted by Coventry at 1:06 PM on December 21, 2007

KirkJobSluder: I was not making any such assumption, because I know better.
posted by Steven C. Den Beste at 2:21 PM on December 21, 2007

Not everyone believes that there is fundamental indeterminacy.
posted by Gyan at 11:00 PM on December 23, 2007