Good post, and I agree with the above. I am saying very little. My only goal is to point out that some people do get overly dogmatic, and some of the replies in this thread were reflexive reactions to this.
This is one of those few scientific areas that plays well in the popular science press. However, the nuances of this particular scientific endeavor often get lost, getting proxied by unjustified ontological language. That’s what one must stay cautious of. As mentioned before, given appropriate caution, theoretical work on quantum mechanical interpretations “is a true scientific endeavor, and plenty of physicists work on it.”
Yes you can; for example it may be be internally inconsistent or ill-defined. In such cases it can be flatly ruled out. But further I would say that there are also indeed cases in which basic philosophical principles such as Occam’s razor can effectively rule out from the table of reasonable discussion observationally equivalent models. One of the big problems is simply that many in physics are not equipped with a deep enough understanding of alternative interpretations of QM to make an educated philosophical judgement. For instance on these very boards there have been pittings of the MWI as being in wanton violation of Occam’s razor, when in fact it is the very paragon of parsimony. The problem is not always derived out of reasonable subjective disagreement, but often, IMO, entirely out of asymmetric understandings of the subject.
If one is ill-defined and another isn’t, then they aren’t equivalent, and their comparison isn’t relevant here. I’m only talking about ruling out one of a pair of equivalent items.
At the risk of making an overly burdensome request, I’d be interested in hearing an example where Occam’s razor can rule out one of two equivalent models.
(I notice that you did put in the qualifying word “effectively”. However, there is an infinite difference between ruling something out and effectively ruling something out.)
Well, as I pointed out upthread, there is at least one class of interpretations that can be ruled out once we have larger quantum computers. But also, I think you are missing the point (at least the point I want to make), which is that these interpretations each can press physics forward in different directions. If one of these directions is more reflective of physical reality, it is likely to yield further insights and models which themselves will hopefully make testable predictions. For example, as I pointed out earlier, Newton’s laws are equivalent to a principal of least action. Should we have scorned the study of the principal of least action? I should hope not! It allows generalization which leads quite naturally to quantum mechanics, which is testable.
The Copenhagen interpretation is ill-defined (what is a measurement?) and logically inconsistent (the measurement apparatus is quantum mechanical). Nonetheless it is functionally equivalent from the standpoint of experiment, and is compared to other interpretations of QM.
To re-emphasize my earlier statement: investigating theoretical interpretations “is a true scientific endeavor, and plenty of physicists work on it.” Theoretical prejudices are the very fuel of theoretical progress, and those with the most highly tuned theoretical intuition have historically been the most successful at advancing our understand of nature.
However, theoretical intuition is fundamentally a tool toward learning – as efficiently as possible – how nature works; it is never a substitute for learning how nature works.
This is so common and deeply taken for granted in physics that it is difficult to step back and point to a single example. For example, the most basic Higgs mechanism for electroweak symmetry breaking. Now, this is not unique. There are an infinite number of more and more grotesque Lagrangians we can construct that are equivalent in their physical predictions. However, anyone plucking one of these Lagrangians out from the morass and, without any motivation whatsoever to overcome the absurd arbitrariness of it, put it forth for consideration, he or she would be laughed out of the field. Because, of course there is a reasonableness criterion here. We aren’t stupid; if the models are equivalent but one is needlessly complex and arbitrary with no motivation whatsoever, it is effectively ruled out. The same way one rules out Russell’s teapot.
The basic problem is that quantum mechanics is so abstract and so dissimilar from our everyday experience of the macroscopic classical world, that assigning an interpretation isn’t easy. We can easily interpret Newton’s laws as we can see them directly in action. Is it surprising that a theory so outside our everyday experience and the way that our brains like to see the world should not lend itself so easily to an interpretation that goes beyond its empirical predictions?
General relativity is another fairly esoteric theory, but at least its basic elements are something that doesn’t need interpretation (i.e. particle trajectories) rather than something as abstract as a quantum state. General relativity can explain everyday situations such as a ball dropping to the floor or the Earth orbiting the Sun simply from it’s first principles. On the other hand quantum mechanics can’t generally explain everyday situations easily as what we think of as everyday physics are emergent properties of complex quantum systems.
Each interpretation has it’s ontological advantage, but also brings with it it’s own problems, for example as great as many world’s theory is as tackling many of the interpretational problems of QM, it’s explanation of observed probabilities is very dissatisfactory. As different interpretations are strong in some areas and weak in others, interpretation can become a matter of personal preference.
People would once have said that about, for instance, the Copernican as opposed to the Tychonic theory of the heavens. They were mathematically equivalent (by design) and so both made exactly the same predictions as to how the planets and stars would move across the sky. Nevertheless, Copernicus was right and Tycho was wrong, and the establishment of that fact was not only a huge scientific advance in itself, but led to many more. Galileo both got himself jailed and founded modern classical physics precisely by refusing to accept that the fact that these theories were mathematically equivalent meant that they were physically equivalent. (And pace Pasta, his achievement very much depended on his conviction that he was right, right that the Earth really is in motion.)
If one construes a theory instrumentalistically, which is essentially what you are doing with QM, treat is as just a mathematical device for generating certain predictions, then, so long as its predictions remain accurate, you have no way of getting beyond it, but if you treat it realistically, which is what various interpretations of QM attempt to do in their different ways, if you take the idea seriously that, say, Bohmian pilot waves, or Everretist alternate worlds might actually exist, you open up the possibility for further discoveries about those things. Real entities have properties beyond those ascribed to them by the theory that originally postulated them.
The current trouble with realistic interpretations of QM is (insofar as I understand it) that we do not have much idea as to what properties (beyond those directly relevant to the math of QM) the entities they postulate might have, or, where we do have some vague clue about that, we have no idea how these properties might actually be observed. That situation could change, however, just as it did for the science of planetary motions, as both the depth of theoretical understanding and, experimental techniques, advanced.
And yet, the Tychonic interpretation is still, even today, used far more often than the Copernican one, and makes the solution of a very large class of problems far more intuitive.
Anyway, this merely reinforces my point. Either way, the math is equivalent and gives the same predictions about what you will see in the sky from Earth (although it may be a bit easier to do the calculations one way or the other). However, despite that, the theories are not physically equivalent, and everybody now knows that Copernicus was right and Tycho was wrong.
With QM we are (perhaps, if we are lucky) in the same sort of position that astronomy was in after Copernicus and Tycho, but before Galileo and Newton showed that if the theories were interpreted realistically they had testable (and wide ranging) physical implications that clearly showed that Copernicus was (in the relevant respect) right. The Jesuit astronomers (and others), before Galileo stirred things up, were quite happy to use Copernicus’ math in an instrumentalist spirit as a predictive tool. If things had been allowed to go on that way, we would never have had the scientific revolution. It was Galileo (who actually cared little for the predictive math of it) treating the Copernican theory as as a physically realistic description of the solar system that was dangerous and revolutionary.
Another instance of observationally equivalent theories where one of which is now near-universally acknowledged to be the ‘right’ one simply because it both yields to a more consistent picture and has allowed further theoretical advances is Lorentz ether theory and special relativity: LET depends on the existence of an unobservable ether to get the right predictions, while SR does not need such an auxiliary construction; additionally, it’s arguable that without the understanding of spacetime provided by SR, general relativity would have been much harder to discover. A purely instrumentalist construal of both theories would have been blind to these (crucial) differences.
The history of science is absolutely full of examples like this. Realism versus instrumentalism about atoms is another example that springs readily to mind. Instrumentalist construals of atomic theory were once prevalent in fields ranging from chemistry to statistical thermodynamics. If they had prevailed, we could not have had stereochemistry (things got nasty over that, I can tell you) or nuclear physics, to name but two fields that make no sense if atoms don’t really exist.
Of course, the trick, and the hard part, is choosing the right things to get realistic about.
It’s typical to envision rational choices driven by something like the expected utility of all potential future outcomes, properly weighted based on likelihood of occurrence, but a person’s preferences might just as easily be described (for an especially strange specimen) as equaling only the single best expected outcome in at least one world, with the others ignored, or even adding a small disutility penalty for all the others where the future self will inevitably realize their world is not the ideal. The calculation is perverse, but the result is clear. There are no unbreakable laws handed down on stone tablets that say a person can’t possibly have the preference of sacrificing the vast majority of their future versions so that the few survivors can make bank. The result can be all the weight given to the most optimum world, with absolutely no marginal weight added to any possible additional worlds.
This sort of person makes different choices based on their understanding of what kind of universe they exist in, regardless of whether the rest of us believe they “should” make these choices.
Reiteration is not an acceptable philosophical (nor scientific) mode, in Western philosophy at least. Nor, in a smaller scale, in long GQ threads. After a while, threads get locked for that.
So far so good.
I don’t know where to begin with this absurd statement.
Huh? As far as I know, that’s the current orthodoxy on the issue: positivism simply doesn’t work (the most easily appreciated aspect of this is the failure of formulating a coherent verificationist philosophy of science, perhaps). Hell, even wiki has this to say on the matter:
[
