- #71
- 8,943
- 2,953
Ken G said:
I think it's mostly BS, and not worth spending time on.
Ken G said:Is it not part of scientific investigation to be able to decide what is a scientific question, and what is not?
And that objection is indeed raised! But note, there is still hope of generating scientific hypotheses using quantum gravity investigations. There is nothing wrong with looking for signposts to the next theory, and interpretations can do that. But note that is still all about the syntax of the interpretation, one simply takes that syntactic structure and extends it. That's what Einstein did to get from SR to GR, he learned some lessons from his own interpretations of the semantic truths represented in experiments like Michelson-Morley, translated them into a syntactic structure, and extended the form of that structure into the arena of gravity. That is all completely within the realm of "what interpretations are supposed to do", but none of that ever had to look like an argument about which interpretation is a correct description of what nature is actually doing. Not only is the latter pure psychology rather than physics, it's worse-- it's a fundamental misunderstanding of what physics does, and what a physics question looks like.atyy said:Sure, but that also means one must object to all research in quantum gravity - it is not science, just interpretation.
I actually don't prefer a belief in a distant reality, I prefer that mode of understanding. I don't actually believe anything is true in some sense outside my understanding of it, I'm quite skeptical that all our modes of understanding are highly restricted by our limitations, but all I mean by the "truth" concept is how I understand things.Sure one can say that. But then your belief in the existence of distant reality is just a belief. You cannot distinguish it experimentally from the nonexistence of distant reality. Yet you preferred one over the other.
atyy said:Read the literature and decide for yourself.
I recommend the discussion in http://arxiv.org/abs/quant-ph/0209123. Weinberg's quantum mechanics book is also very good https://www.amazon.com/dp/1107028728/?tag=pfamazon01-20. I think most will agree with those two recommendations. The one contentious recommendation is Ballentine, which most people like, but I think is somewhere between misleading and wrong.
atyy said:Oh yes, and before you read those - read Landau and Lifshitz https://www.amazon.com/dp/0750635398/?tag=pfamazon01-20 - that's a real physics book - it does not shy away from interpretation, it states a practical view and its limitations clearly, as a basis for physics.
After Weinberg that you can also read Haag https://www.amazon.com/dp/3540610499/?tag=pfamazon01-20 - another real physics book that shows that interpretation is central to the thinking of great physicists.
I agree that scientists are also people, but they have a "scientist hat" they put on when they wish to be regarded as a scientific authority, and a "person hat" they put on when they just want to share their beliefs with everyone else. This is quite important, because when these hats get mixed up, we lose the authority of science and just become two people spouting their own faith systems. Of course people like creationists will always want to turn the debate into something of that form, but the scientist must not fall for the trick-- the scientist is only a scientist when they are actually doing science, and they only have a right to speak in a science classroom, or science TV show, etc., when they are wearing the right hat.stevendaryl said:I think that's BS. Scientists are just people. They are just ordinary people. They believe things, sometimes for good reasons, and sometimes for silly reason. There's nothing that scientists are required or forbidden to do.
I never said what you can think about, and I certainly never said you shouldn't think about interpretations! I said you should understand what the thought process is that is actually happening there when you do think about interpretations. You should understand that the thought process is "I gain understanding A of experimentally verified semantic truth B by contemplating syntactic structure C." That's understanding what an interpretation is, and what science is, and you get everything that every scientist ever got-- and you never needed to say "and I can only get these things if I believe C is what nature is really doing."You could say that science is a certain way of exploring nature through formulating and testing hypotheses. But to formulate a good theory or to come up with a good test for an existing theory requires thinking about the theory. It's way too rigid to say: "When you're doing science, you can only think about certain things."
It is the scientists who need to understand their own thoughts, they need to know when they are not doing science. It's fine if it is the scientists who decide this-- your model is only a problem if it comes from outside science, from people who do not understand science. But when scientists themselves lose track of what doing science actually is, this is a problem, and the way "the measurement problem" is sometimes framed is actually this problem.I think you have to be careful about creating thought police to inflict on scientists.
Ken G said:And that objection is indeed raised! But note, there is still hope of generating scientific hypotheses using quantum gravity investigations. There is nothing wrong with looking for signposts to the next theory, and interpretations can do that. But note that is still all about the syntax of the interpretation, one simply takes that syntactic structure and extends it. That's what Einstein did to get from SR to GR, he learned some lessons from his own interpretations of the semantic truths represented in experiments like Michelson-Morley, translated them into a syntactic structure, and extended the form of that structure into the arena of gravity. That is all completely within the realm of "what interpretations are supposed to do", but none of that ever had to look like an argument about which interpretation is a correct description of what nature is actually doing. Not only is the latter pure psychology rather than physics, it's worse-- it's a fundamental misunderstanding of what physics does, and what a physics question looks like.
So I say more power to anyone who presents an argument like "if we interpret QM through the eyes of interpretation X, it suggests we do experiment Y, because we might get surprised by that outcome." Like Bell did. So we get "any interpretation that looks like Z doesn't work for experiment Y", but someone is then free to modify interpretation Z into Z' such that it still works. That's what Bohm did. That's all good stuff, it's using interpretations the way they are meant to be used, no debate around which one is what reality is actually doing because that's not even a physics discussion any more. The physicist should not expect his/her interpretations to represent their own semantic truths about reality, they should regard them as syntactic structures that can be used to derive semantic truths that are verifiable by experiment.
What I'm saying is that principles of physics theories are not semantic truths. So conservation of energy is not a semantic truth, it is a synactic structure that let's us predict the semantic truths we test. The idea that all observers will get the same result for the speed of light is not a semantic truth, it is a syntactic structure that let's us derive the semantic truth that is the Michelson-Morley experiment. But we are free to do an experiment tomorrow that violates either of those principles, all it means is that we need a new syntactic structure to get our semantic truths. So has it always been, throughout the history of science.
The "we" concept is a complex notion, not a simple semantic truth or syntactic structure, and not strictly a physics notion either-- nor do I invoke any scientific authority when I invoke that term, indeed I know no more about that term than a person who has never sat in a physics class in their life. There are many things we do not have physics for, that's one of the reasons that the physicist is not yet in the physics. We use semantic truths and syntactic structures when we do physics, but that doesn't mean we don't also invoke more complex notions when we need them. We couldn't even get out of bed in the morning if all we had to go on was physics!atyy said:When you say "But we are free to do an experiment tomorrow that violates either of those principles, all it means is that we need a new syntactic structure to get our semantic truths. So has it always been, throughout the history of science." is the existence of the "we" that you refer to a semantic truth or syntactic structure?
A scientist can have hope, when did I say they cannot? Their hope is not their science, however-- they have a hope their approach will lead them to a scientific result that will not be just another footnote to history. But whether it will or not will be tested by science, not by hope or belief.Also, I don't think you are consistent on the issue of quantum gravity. First you say that being testable in principle is not enough, then you say that there is still hope for generating testable hypotheses from quantum gravity. Presumably you mean the hope is not scientific, it's just a belief and those who investigate that hope are not scientists?
Yes, I have no issue with using an interpretation to guide scientific investigations, and I am quite sympathetic to the idea that all physics theories are "effective" theories.At any rate, here is some work using two realist "interpretations" of quantum mechanics to make predictions - I put "interpretation" in quotes because these both say that quantum mechanics is probably only an effective theory, and will accordingly fail somewhere, just as GR probably will fail. I say that GR will "probably" fail because it has not been ruled out that GR is Asymptotically Safe, just as we have not ruled out Many-Worlds as a coherent possibility.
http://arxiv.org/abs/0805.0163
http://arxiv.org/abs/1306.1579
http://arxiv.org/abs/1405.2868
Ken G said:Yes, I have no issue with using an interpretation to guide scientific investigations, and I am quite sympathetic to the idea that all physics theories are "effective" theories.
True, but that problem comes with the territory of science, so in some sense it's not a problem-- it's just science. Sure we always want better science, so we have a "problem", but our problem is not that our theories are effective theories, our problem is that we have more to learn. But isn't that why we do science in the first place? So there's nothing wrong when science has more to learn.atyy said:What I don't get about your position is: if quantum mechanics is an effective theory, then there is a problem - quantum mechanics is incomplete.
I'm saying we shouldn't even try to understand quantum mechanics, or any physics theory, as a "complete theory." That is not something we test, it's not an attribute of any theory. It's certainly not a semantic truth, because it is not testable, and it cannot be part of the syntactic structure of any theory, any more than there can be an axiom of arithmetic that says arithmetic is complete. Indeed, as I'm sure you know, Godel proved that putting an axiom like that into arithmetic is guaranteed to make it inconsistent. What I want to know is, from where comes this fascination to regard any physics theory as a complete theory? None has ever been, yet countless times the proponents of various theories expected them to be. I think at some point we step back and learn our lesson as to what physics really does, and what it does not do.The measurement problem asks - can we understand quantum mechanics now, in its present form with a Heisenberg cut as a complete theory, or the most complete theory possible? The Bohmian interpretation, and other realist interpretations are that we cannot understand quantum mechanics as a complete theory, but we can understand it as an effective theory. Are you saying that we can understand it as a complete theory, or the most complete possible, and there is possibly no effective theory underlying it?
There is no measurement problem because quantum mechanics is an effective theory. The "measurement problem" is the question "which interpretation is what nature actually does, such that quantum mechanics plus that interpretation is a complete description of reality." That's not a scientific problem, because a proper framing of science is not about that, it's about creating effective theories like all the other theories of science.That is coherent, but I think you asserted both that there is no measurement problem and quantum mechanics is an effective theory. Why would one think quantum mechanics is an effective theory if there is no problem to be solved?
Ken G said:There is no measurement problem because quantum mechanics is an effective theory.
Ken G said:That's hard to answer, because it is basically asking if quantum mechanics wasn't quantum mechanics, and physics wasn't physics, and humans weren't humans, would there be a measurement problem? But I think you are asking if quantum mechanics was what nature is actually doing, and one of its interpretations was actually the complete truth, would there be a measurement problem? I would have to say "no" to that, because the "true interpretation" has done away with the apparent problem of the "Heisenberg cut." In that case, we would have eliminated the Copenhagen and ensemble interpretations, because those interpretation more or less assert that quantum mechanics is by its nature an effective theory (in different ways, but both say it's effective because it doesn't say how collapse happens after decoherence has set the stage). If Many-Worlds were "true", then there's no measurement problem because it is the nature of reality that each consciousness cannot know all of it (so quantum mechanics is not an effective theory, but scientists are ineffective witnesses). If Bohm were "true", then there's no measurement problem because the "classical world" we imagined pre-quantum is actually quite right, but there are additional "spooky" elements to it that we did not appreciate before. But I'm saying that the reason there is no measurement problem is because it is not actually a "problem" that quantum mechanics is an effective theory, as "effective theory" is redundant word use.
Ken G said:OK, so if you agree that "effective theory" is redundant, then you must feel the value of Bohmian mechanics is not that it allows us to reframe quantum mechanics as a complete theory of nature, but rather simply that it provides a signpost for a direction to a new theory that might one day be confirmed to make different predictions than quantum mechanics? In that case, there's no problem, because it will be those future observations that decide the issue, and the only discussion we could have now is what are the right observations to do to test this hypothesis.
I agree with that, I just don't think it matters if there is a theory that quantum mechanics is an effective theory of-- because physics theories are always effective theories. But that only means that your point has no traction with me, it certainly would have traction with anyone who thinks quantum mechanics is like a "dead end" sign. I just can't see that view, why would this one theory, out of the long list in our advancing understanding, be the dead end? I tend to think the differences between quantum mechanics and classical physics are highly overblown, as would a Bohmian, but I may think that for a different reason: I think classical physics had the same "problems that are not really problems unless we take the theory too literally" that quantum mechanics does, whereas a Bohmian tends to think neither one has a problem being taken literally.atyy said:Yes, but not just that - if someone says that quantum mechanics is a fundamentally new type of theory, and that the "true interpretation" must be the Copenhagen interpretation, because it is impossible to construct any theory of which quantum mechanics is an effective theory - then theories like Bohmian Mechanics show that that need not be true. In other words, I can actually back up my statement that "The measurement problem can be solved by considering quantum mechanics as an effective theory" by providing examples of theories of which quantum mechanics might be an effective theory.
Yes, I agree with that basic dichotomy, though I think it's odd that many people see the larger problem as the second type! Instead, I see that second type of "bug" as a feature. A theory that gives you no clue when it breaks down just beguiles you into thinking it might not be an effective theory, but of course it is, and gives you no guidance for how to explore that. A theory that gives you a clear signpost as to where it is difficult to regard as anything but an effective theory is a good thing, because it does give guidance to the next theory. What's more, it is not necessary to frame the search for the next theory as a search for the complete theory, it is only necessary to build a syntactic scaffolding around the semantic truths that are essentially "rules of thumb" in the one theory (and all theories have these, they are the places you can ask "but why is..."), such that you can derive those semantic truths from the syntax of the interpretation. Then you extend that syntax to new semantic truths that have not been observed yet, and voila, there's your guidance to the new theory (general relativity is a perfect example of this approach). At no point does any of that have to look for a search for a theory that is not an effective theory, and I would call it an error to frame it as such.For me, I would prefer to say the measurement problem exists and has been at least partially solved, rather than that it does not exist. The reason is that even if all theories are effective theories, they divide (as a matter of practice) into two types. The first type of effective theory does not reveal any obvious incompleteness in itself (even if you use Goedel's incompleteness theorem, a theory that is subject to it need not be obviously incomplete, because all Goedel sentences are of the form "for all ...", whereas an experiment usually tests a statement of the form "there exists ..."). So usually we only know how to falsify this first type of effective theory by experiment. The second type of effective theory reveals an obvious incompleteness in itself, and provides a theoretical opportunity - a signpost to new physics - as you say.
I agree, but I take that as my default stance about all theories, so I don't need the theory to tell me it is of the second type. All that matters is whether the theory is giving us guidance as to where to look for how to advance to the next theory, and it's hard to know that until we try. Winners write the history on this! If the Bohmian approach provides true guidance to the next theory, everyone will say it was obvious that quantum mechanics had this Heisenberg cut problem, and they eventually got the better theory by rejecting that cut. But if the Bohmian predictions don't work out, it will be forgotten that anyone ever suggested they would!Quantum general relativity (if it is not asymptotically safe) is one such theory. Newtonian gravity and Maxwell's equations were another such theory. And I think the measurement problem indicates that quantum mechanics also belongs in this second class of effective theories which signal their own incompleteness, even before any experiments have falsified them.
Yes, and if that remains the "end of the story", it would be even worse than if Bohmian mechanics turned out to be right! That would be the worst case of all-- for the present state of affairs, where we have many worlds and multiverses and machinery that is flexible enough to explain anything you want without any chance of ruling it out, to continue indefinitely.In both quantum general relativity and quantum mechanics, since we have not ruled out that quantum general relativity is asymptotically safe nor have we ruled out Many-Worlds as a possible interpretation, it remains possible that they are effective theories of the first type.