The conventional wisdom among people who know a little bit of quantum mechanics is that quantum mechanics is weird.
The conventional wisdom is wrong. Quantum mechanics is not weird. Interpretations of quantum mechanics are weird.
My thinking on this has changed over the years. In high school I read everything I could about the “weirdness” of our universe: Schrödinger’s cat, wave-particle duality, the collapse of the wave function, many-worlds theory, the Heisenberg uncertainty principle.
Then a strange thing happened: I went to college. I studied physics. And guess what? None of that stuff gets more than the briefest mention in the physics classroom. Why?
Because those things are beside the point. Quantum mechanics works. How you interpret quantum mechanics is your problem.
There’s a dichotomy here which is the source of most people’s confusion. Theories are different from interpretations of theories. A theory is a mathematical model that allows us to make predictions. An interpretation is a philosophical construct that allows us to sleep at night; it is a squishy heuristic that helps us unimaginative humans make sense of the math before us. Theories get things done. Interpretations never helped anybody, not really.
Let’s say that in an abandoned shack you discovered a notebook with the word “PHYSICS” written by hand over and over, thousands of times, apparently filling every page. You haven’t looked at the last few pages, but your theory is that these pages will also have the word “PHYSICS” written out. Each time you turn a page, your theory is validated: “PHYSICS” is there, as predicted.
Next to this notebook is another that looks just like it. You open the first page, and are not surprised to see “PHYSICS PHYSICS PHYSICS” again. What’s going on? Did some crazy person live in this shack? Such speculation doesn’t really matter, since you can still hypothesize that “PHYSICS” fills this notebook as well. In fact, you have a stronger theory: every notebook in this shack is filled with “PHYSICS”.
You perform an experiment: you turn the page. “PHYSICS PHYSICS PHYSICS”. The experiment supports your theory. You find more notebooks; same results. Every notebook in the shack is filled, apparently, with “PHYSICS”. But guess what? There are dozens of possible interpretations. And in the absence of further data, you can never know which one is “correct”.
Maybe the shack was once inhabited by a crazy person, who wrote “PHYSICS” precisely 250,001 times in a futile attempt at summoning Cthulhu from his ancient slumber.
Maybe a student misspelled “physics” on a test, and her cruel teacher punished her in the most depraved way possible.
Maybe Matt Damon filled the notebooks, in a method-acting attempt to get into the mindset of an OCD scientist.
Which of these interpretations is the “truth”? Without further data you cannot really say. Arguing about which is right and which is wrong is futile at best, and annoying at worst.
Of course, new data may turn up. We might find out that the notebooks are 75 years old, ruling out our Matt Damon interpretation. That interpretation is no longer a valid interpretation of the data.
Which brings me to my next point: there is no official arbiter of what constitutes a theory versus what constitutes an interpretation. Different philosophers and scientists have used the words differently at different times. All you can hope for is that a particular author is consistent in his/her use of the terms. I personally use the word “interpretation” to describe competing theories that cannot currently be differentiated by any known scientific experiment. If two different interpretations make different, testable predictions, then they are promoted to being totally different theories. (Caveat: others use the words slightly differently. Deal with it.)
So what does this have to do with quantum mechanics?
Quantum mechanics is an entirely mathematical theory. Its postulates are logical, concise, and powerful. We can use quantum mechanics to invent cell phones, computers, lasers, and iPods. Quantum mechanics doesn’t care if you “understand what it really means”, or not. It is arguably the most successful and powerful theory to come out of the 20th century.
Now, the mathematics of quantum mechanics are abstract and hard to visualize. Nevertheless, people insist on trying to visualize anyway. And the result is all kinds of weirdness: Schrödinger’s cat, wave-particle duality, the collapse of the wave function, many-worlds theory, the Heisenberg uncertainty principle. These ideas are all mental hoops that people have jumped through to explain some unambiguous, concrete, abstract linear algebra. The math is just math, and it works; what it means is anyone’s guess.
There’s no crying in baseball, and there’s no philosophy in quantum mechanics.
Don’t like the many-worlds interpretation? Fine. Be a Copenhagenist. Don’t like pilot waves? Great. Stick to your pet idea about superluminal communication. Just remember that all of these competing interpretations make the exact same predictions, so for all practical purposes they are the same. Some people go so far as to say, just shut up and calculate. [Note added 3-19-14: there are problems with pilot wave theories that in my view rule them out as being a valid interpretations of quantum mechanics. But there are hoops people can jump through to try and “force” pilot wave theories to be consistent with, say, Bell’s theorem. My broader point is that there are multiple interpretations of QM and that all have followers to this day, but that none of the interpretations really have any distinct implications for our lives.]
I don’t usually go that far. I actually think that the many-worlds interpretation is a testable theory, not an interpretation (hence the name of this blog). I think many-worlds is falsifiable. (If we ever observe a wave function collapsing, then many-worlds will have to be discarded.) But I don’t think that will happen: many-worlds is too elegant, and too powerful, to not be true.
But we’ll see.
If you think it’s absurd that a cat can be alive and dead at the same time…if you think that it’s crazy to hypothesize other universes…if you think that God does not play dice with the universe…don’t blame quantum mechanics. Blame the philosophers who try to interpret it.
Quantum mechanics works. Otherwise, you’d be reading this on an actual piece of paper.
Curious if you’ve read David Deutsch (e.g., The Beginning of Infinity). His treatment of these questions contradict yours to some degree and is more convincing. (He is pro- many-worlds.)
You seem to dismiss worrying about so-called interpretations — saying “Just remember that all of these competing interpretations [of quantum mechanics] make the exact same predictions” — but then you say many worlds (which is typically considered an interpretation) does in fact make different predictions. So do think it’s worth worrying about or not? Is many worlds theory not part of quantum mechanics?
When you say quantum mechanics is an “entirely mathematical” theory, I think you are referring to the significant mathematical machinery that allows predictions separate from a complete theory. But to me that only means that part of quantum mechanics (and all the rest of the current theory) is so far only an incomplete theory. Indeed it doesn’t even make sense to say any physical theory is truly “entirely mathematical” — how do you link the mathematical symbols to the physical world? (Answer: it’s part of theory.)
Lastly, in your “PHYSICS notebook” example, you imply it’s pointless to consider theories… unless data “turns up.” Well, you don’t really have data until you have a theory in the first place. There is an infinite amount of possible empirical observations about the notebooks (e.g., the absorption cross section of the ink over all wavelengths as a function of time of all time), but only the subset of it that affects a plausible theory (or explanation, to use Deutsch’s term) would be considered data.
1) I think that the MWI *should* be separated out from the other interpretations…I think it *is* superior, for many reasons…
2) QM is entirely mathematical, and I think ultimately any theory of everything will be mathematical only. (I ascribe to Tegmark’s MUH….mathematical universe hypothesis). I believe that if you can’t quantify something in terms of mathematics, then you’re just whistling Dixie.
3) I am probably using the word “data” differently than you. To me, data is any observation I make. I can make observations without any preconceived notions. The job of a scientist is to take that data then think of a plausible theory that fits the data. I’m not sure why you have to have a “theory” in the first place to make the observation. If I walk into a shack and see a notebook, I have some data now, even if I haven’t thought at all about possible explanations.
Matthew –
1) So I guess you don’t think MWI is part of quantum mechanics (since it’s not “entirely mathematical”)? Or maybe (from your last addendum) because it doesn’t have “implications for our lives”? (It has to be practical to be physics? Ha.) But yet you say it’s falsifiable, making it part of science in the Popper sense. What is the value of separating it out like that? Again, I favor the Deutsch term that science is finding good “explanations,” and both the math and the interpretation seem like fundamental parts of that.
2) Now I at least know where you’re coming from with the “entirely” mathematical part, when you mention Tegmark’s new book: http://space.mit.edu/home/tegmark/mathematical.html I’ll have to read that. And I see Deutsch has blurbed it too. (I will note that Tegmark himself labels the idea “extremely controversial.”)
3) I think it’s the word “theory” that we are using differently — I would include common sense theories, a necessary part of any complete explanation. (I’m not suggesting you need a complete explanation before you can have data!) That said, I know what you mean, I think, about approaching data without “preconceived notions.” I’ve heard other scientists explicitly say it’s *bad* to approach data with possible theories already in mind. This is fine, I think, if interpreted as: try to “avoid confirmation bias,” (keep an open mind). But if taken much further I’m not sure it’s even coherent. (Popper mocked the Baconian observationalism you seem to hold by telling his class: “Observe” and waiting for someone to ask “Observe what?” ) The scientific process, as I see it, involves making conjectures/theories (which have priori probabilities, using a Bayesian framework) and using observations to judge between them (updating the prior conjecture with data). You wouldn’t start “taking data” on the precise position of every atom in the shack because no plausible and relevant conjecture depends on that type of observation. Similarly, you might as well use more plausible conjectures to motivate future data-taking. So arguing about interpretations without data is only “futile” (your term) *if it never motivates the search for new data/tests*. But why would you assume that? That’s the heart of science.
I was also going to make a comment about Deutsch’s take on this but you’ve covered it. All I’d add is that Matthew’s point 2) is taking a philosophical viewpoint, just a bit further upstream, so that seems to invalidate the argument that you shouldn’t worry about the interpretations at all. If you really want to argue that you shouldn’t be philosophical about quantum mechanics, then you don’t get to argue that you shouldn’t be philosophical about quantum mechanics as that is a philosophical stance.
I am not making myself clear. I am not arguing that we shouldn’t do philosophy…I enjoy it, and do it myself. Nor do I argue that I myself don’t have a philosophical stance, My thesis is this: don’t call QM in and of itself weird; it’s just some equations that work. If you want to dismiss QM and call it weird and call it BS and shake your head at how bizarre modern physics is, then blame the philosophers, not the physicists.
“Maybe the shack was once inhabited by a crazy person, who wrote “PHYSICS” precisely 250,001 times in a futile attempt at summoning Cthulhu from his ancient slumber”
LOL! If you ever get tired of Physics you can always be a staff writer for the Onion 🙂
If you know someone at the Onion, put a plug in for me, will you?
Wish i did.
From THE AMAZING STORY OF QUANTUM MECHANICS: One of the most amazing aspects of quantum mechanics is that one can use it correctly and productively – even if one is confused by it.
I am confused about how my car works as well. But it still gets me to work on time!
I suspect that the theory-interpretation distinction wasn’t often made before quantum physics came along. Most earlier theories were pretty easy to interpret. Heliocentrism, for example, was counterintuitive at the time, but not hard to interpret. The germ theory of disease was actually easier to interpret than the mischmasch of earlier theories. Relativity was difficult, but not impossible. But quantum theory is impossible.
Schrodinger’s cat: why is a human observer dragged in? Isn’t the paradox resolved the instant any event registers it in the macro world? The sequence I’ve read is string is cut / hammer hits poison vial / cat is poisoned / human sees cat. Why do you need the later steps? The resolution came when he string was cut.
A lot of German physicists were bizarrely idealistic in the metaphysical sense, in a way only loosely connected to their scientific work.
I don’t think we were discussing Schrodinger’s cat at all. Anyway, no one really believes in the literal superposition of states (that I know of).
You just happened to mention it.
I think that Schrodinger’s cat was a thought experiment to “prove” that the idea of superposition is absurd. But with the theory of decoherence on pretty firm footing now, no one but popular science pundits even mention it. Oops, and people like me. Sorry.
By the way, is that Mussorgsky?
Yes, I’m a big admirer.
That makes me feel better about Schrodinger. It always seemed goofy to me, but people took it very seriously.
The thing is that Bohr’s philosophical stance in the Copenhagen interpretation, whilst drawing implicitly on a tradition of heavy-duty epistemology of scientific observation – which wasn’t unusual for continental physicists of his generation and background – was precisely to take philosophy out of the physics. We as humans can’t help but force interpretations on things. It is what our brains have to do to make sense of, process and navigate the world. And more importantly, science before the 19th century made a lot of headway intellectually because it was capable of interpreting and rationalising many of our everyday concepts, and the ‘weirdness’ of QM is more a consequence that we don’t have ‘common-sensical’ pictures of the concepts it describes to latch onto. Analogies, by always granting some kind of classical crutch, always make it seem weirder because our image of it is not of a coherent logical and observation theory of physical laws, but as a sort of deliberate ‘breaking’ of a classical picture.
I mean our understanding or thinking of anything is essentially conceptual (even pictorial), for good neurological reasons and at a certain level of the mathematical understanding of the universe that is simply not congruent with the sort of intuition we often want. We can hold interpretations or conceptualisation which are mutually incoherent (particle-wave duality), simply because of the way our mind processes the world around us. And that includes the moment of measurement which seems so mysterious. So to that extent CI isn’t even a philosophical interpretation to be held, but an explanation of why there ARE competing philosophical interpretations, and an attempt to strip away as so much mental interposition all but what is necessary for us to simply interpret what we’re doing as calculations without positing more. And indeed, it is why the spirit of CI would live on even if MW were to be proved.
In “The Fabric of Reality” David Deutsch relates a Bertrand Russell tale about an intelligent chicken in a barnyard who notes that the farmer comes in and gives the chicken food every day. The chicken eventually hypothesizes that the farmer will continue to do this forever, based on past observations, and also hypothesizes that the farmer is a benevolent being who likes chickens because he feeds them every day. One day the hypothesis is falsified when the farmer comes in and wrings the chickens’ neck. If I understand Deutsch correctly, he is using this as an example of the failure of the logic of the inductive process, as you are using the crazed OCD scientist (or whatever) writing “physics” over and over. Deutsch says that scientific theories should “explain” observations, not merely record them, a point of view that is contrary to much of modern science, although, now apparently everyone agrees that any valid theory must be “falsifiable”. The many worlds “interpretation” is an explanation, “shut up and calculate” is not, although both may be falsifiable. I took 2 semesters of college physics 40 years ago, have I gone down a rabbit hole?
When one says there is no philosophy in something, they are making a philosophical assertion. The Positivist hangover continues in not separating pure math from practical. An instrumentalist view of science, (it works), is good as far it goes. If nothing else, philosophy can help make sure reason drives decisions on what to experiment with, instead of social forces and pet theories of certain groups.
How does apply in any way to anything I said? I just don’t get it.
“unambiguous, concrete, abstract linear algebra”
dang!
I don’t see how you can say interpretations are not important. They tell us what we are looking at, what questions to ask, and more importantly what’s really out there. Mathematical theories are nice, but they’re static and empty; there’s no progress, nothing to wonder about, nothing to drive us further to the next scientific discovery. Look at Ptolemy and his theory of planetary motion: it perfectly describe the location of the planets in the sky, mathematically calculated and predicted their exact paths, and to this day we could still use his theory to find where Mars might be, but it was wrong. Ptolemy had no idea what the true motion of the planets were, and his mathematical theory offered no enlightenment. It is the philosophers, the interpreters, that push science forward even when something has “for all practical purposes” been worked out.
Apparently I cannot express myself at all about this. I do not mean to say that philosophy is unimportant, or uninteresting. I only mean to say that *disagreeing with an interpretation on philosophical grounds* is not a valid reason for rejecting QM. That’s it. Since there is (currently) no interpretations better than Copenhagen or Many Worlds (really, the only two interpretations that fit all the data) you cannot reject QM if you don’t like those interpretations. And as for actually *doing* QM, we have 85 years of experience that say that interpretations are as useful for day-to-day work as tits on a boar.
I agree that “*disagreeing with an interpretation on philosophical grounds* is not a valid reason for rejecting QM”. I was objecting, and maintain my objection, to your statements such as “Interpretations never helped anybody, not really” and “Arguing about which is right and which is wrong is futile at best, and annoying at worst.”
OK sure.
I will happily bet an arbitrary 6 pack of your preferred beer against the proposition that “Theories get things done.”
I don’t understand. You don’t think theories are useful? I find it hard to see how we could have gotten to the Moon without Newton’s Theory of Gravitation, for example.
Whatever.
Hi,
Somehow, you made a philosophy interpretation about QM. Your point of view is a subgroup of Copenhagen interpretation.
Also, according to Heisenberg, the language (interacts with clasdical logic directly) has limitations to interpret QM’s mathematics.
It is very difficult for us to avoid certainty.
Good luck
Subgroup of Copenhagen? Nope. And “the language has limitations…” This is gibberish.
Matthew, you don’t seem to realize how much your reasoning is philosophical.
A profoundly misguided notion others have pointed out is that “A theory is a mathematical model”. While some theories could be categorized this way (with loads of caveats), most of what we consider theories would not be properly categorized as “mathematical”. Maths are more the languages of models within theories, AFAIK.
Yet this process of categorization (in which you engage here) is philosophy of science, philosophy of physics, or philosophy of math, *not* physics or math.
Physics can be no more free from philosophy than baseball can be from physics (Prof. Jeff Kasser)
From a very narrow view, physics has no relation to baseball, but we would rightfully consider a Yankees team member making such a claim to be quite naive about his profession and how it works – no matter how good a player he is. We can simply observe he should stick to hitting homers.
When you claim: “Since there is (currently) no interpretations better than Copenhagen or Many Worlds…” you are assuming a metric by which one should evaluate competing theories. Standards for theory evaluation are not part of physics, they’re philosophy. You are attempting to deploy tools about which you have no knowledge, and the reliability of the resulting cognitive structure is about what anyone should expect: pretty bad.
After a naive, undefined selection of meta-physical rules, you make another philosophical claim (normative this time) that: “you cannot reject QM if you don’t like those interpretations.” On the contrary, there are a host of reasons why one might not like interpretations that are entirely legitimate on philosophical grounds, such as pragmatic rules: that such interpretations, theories, and assumptions have proven disastrous to the advance of human knowledge, for example.
Finally, the fact that some expression is unintelligible to me is no reason to label it gibberish, unless I’m simply determined not to consider what it might represent – in which case such uninformed dismissal efficiently protects my willful ignorance, but I do tend to excuse myself from reasoned, adult discussions by such claims.
Frankly, I don’t know what you’re arguing about, or what your point is. I still maintain that the sentence “Also, according to Heisenberg, the language (interacts with clasdical logic directly) has limitations to interpret QM’s mathematics” is gibberish. And I stand by my statement that QM is just a mathematical theory, which succeeds or fails based on whether or not it models reality in a useful way. How you interpret it is philosophy. But the utility of QM in no way depends on philosophy.
Your car – exactly like Quantum Physics – works just fine until it doesn’t work any more… and any sensible physiscist can tell you that QP does not ‘work’.
What you learned in a low level undergrad dive into QP completely ignores the tough questions – ie. exactly how and why does a wave collapse – this is still being taught like it was 1935.
If you are not intellectually stimulated by the deeper questions of QP then you should have been an engineer (which I was).
Hello Matthew, After James’ comment, I just realized I hadn’t responded to your reply that you didn’t understand my argument. The view from my profession is that unless one is trained in information system analysis and risk management, that person is not a possible source of reliable characterizations for the theories *within* which they may be experts. They are generally not qualified to assess structural features, relative prioritization, and broader reliability or applicability of the field outside their experience.
While you are clear and precise on applicability and deployment of QM for specific problems, your definitions for QM appear poor to one who has worked on industry standard glossaries and lexicons. We would expect this from a non-specialist taking a first crack at creating definitions. Good ones are difficult, take a long time, and require deep understanding of use and the context of use.
Absent these inputs, one is likely to make claims like “baseball in no way depends on physics”, or “interpretation is philosophy”, etc. The decision to use a tool is different than the tool.
Clear reasoning about relations and importance is like describing relativistic motion: we must specify our frame of reference. When we say “QM”, are we referring to the phenomena that give rise to human observation? …the body of concepts and theories that fit under that term? utilization or interpretation of one of more of those theories?
To discuss productively, we need to understand others and be understood by them. For that, we need to provide evidence for our criticism. Stating a claim (e.g.: Heisenberg’s) as “gibberish” could appear to readers as indistinguishable from an “argument from ignorance” or “argument from personal incredulity” unless we identify what the claimant meant, and where/how the claim goes wrong.
I believe “language has limitations to interpret QM’s mathematics” is personally intelligible, and I tend to agree with it as I do with Wittgenstein’s assertion that “limits of my language mean the limits of my world”. If my interpretation and agreement are misguided and the claims truly are “gibberish” as you say, I would like to understand how I might change my opinion.
Is the Bayesian reading of QT an “interpretation” in your sense, and thus outside QT itself? Its my experience that Bayesians in a range of scientific fields tend to believe that they are both contributing to those scientific fields AND doing philosophy, so I suspect Fuchs and company would disagree with what you seem to see as a dichotomy.