9 ms·

Measurements on the reality of the wavefunction

36 points by MichaelAO 12y ago

cma 12y ago
Full preprint article: http://arxiv.org/abs/1412.6213 http://arxiv.org/abs/1412.6213
- jriordan 12y ago
  Okay, I'm like 5 years old, but could you explain to me how this is available here for free, yet it costs $32 to read the final published version? I mean... this is fundamental human knowledge we are talking about... isn't there some better way of doing this than $32-a-pop?
  - chriskanan 12y ago
    I have institutional access so I can't easily check, but I thought most or all Nature journals were now freely available: http://www.nature.com/news/nature-promotes-read-only-sharing-by-subscribers-1.16460 http://www.nature.com/news/nature-promotes-read-only-sharing...
    - jsprogrammer 12y ago
      From the link you posted: >All research papers from Nature will be made free to read in a proprietary screen-view format that can be annotated but not copied, printed or downloaded... >Annette Thomas, chief executive of Macmillan Science and Education, says that under the policy, subscribers can share any paper they have access to through a link to a read-only version of the paper’s PDF that can be viewed through a web browser. The person who linked the original article didn't use the special link to the "free, but proprietary" format.
  - jsprogrammer 12y ago
    > but could you explain to me how this is available here for free, yet it costs $32 to read the final published version? Nature put a paywall on top of the article and then someone published the paywalled link on HN.
    - TTPrograms 12y ago
      I don't think that's quite right. Arxiv is a third party for academics to publish paper to get informal peer review outside of journals. My guess is that any edits made based on feedback from Nature's author network as well as the actual paper in the Nature format are under Nature's copyright. But yes, my understanding is that anybody with institution access to Nature can now share a link to the Nature publication.
      - jsprogrammer 12y ago
        Arxiv shows that the most recent revision of their document was published on January 20, 2015 (less than three weeks ago). I suppose it is possible that Nature's version is substantially different, but if that were the case it would cast a huge shadow of doubt over Nature's editorial process, as well as the claims made in the paper itself.
        TTPrograms 12y ago
        The differences are likely a few small sentence additions, citations and comparisons to other work. I think the results would remain unaltered.
        montecarl 12y ago
        Peer review and getting an article published takes at least a month if not half a year or longer. The preprint on Arxiv is most likely the final version of the paper they submitted for publication post peer review.
      - privong 12y ago
        > My guess is that any edits made based on feedback from Nature's author network as well as the actual paper in the Nature format are under Nature's copyright. Half correct. The authors retain copyright for changes due to feedback (peer review), as the authors themselves would have modified the paper to address any feedback. You are correct that the actual paper in the Nature formatting (and possibly associated copy-editing) are under Nature's copyright.
TTPrograms 12y ago
Every now and then people (on HN and elsewhere) like to rediscover pilot-wave theory and other similar ideas that posit that billiard-ball-like-particles still describe the universe. In these interpretations the wave-function, which is the probabilistic field that modern physics uses to model particles, is just an effective representation about our knowledge about the system. I tend to see these theories as paying a price of drastic increase in complexity to retain classical intuition about the system. The mainstream view in modern physics (for the past 30 or so years, I believe) is rather that the wavefunction is the fundamental description of reality, at least in the sense that classical effective knowledge theories are not plausible models. My interpretation of this paper on quick glance is that it is an attempt to experimentally demonstrate that only effective knowledge theories with very strange properties would suffice to explain observations, thus strengthening the "wave-function is fundamental" viewpoint.
- jriordan 12y ago
  What do you think is the best way to learn of these things, i.e. physics fundamentals and all in the modern state of knowledge? I ask from the perspective of a B.S. Physics cum computer software person... which, I think, might not be too far removed from the perspectives of most HN peeps...
  - TTPrograms 12y ago
    Understanding current thought in the field for me is based on skimming many academic papers. That's kind of different, though. Getting a solid understanding of the fundamentals of the various fields is more of a coursework sort of thing. There's a bunch of quantum lectures on MIT's open courseware - anything Allen Adams will be great. If you're a book kind of person I believe I enjoyed Shankar but it's been a while.
- phkahler 12y ago
  >> Every now and then people (on HN and elsewhere) like to rediscover pilot-wave theory and other similar ideas that posit that billiard-ball-like-particles still describe the universe. Have you seen the experiments of Yves Couder? Hypothetically, if a model like that produced an exact mathematical equivalent of quantum theory wouldn't that be more satisfying? It would also destroy some of the notions that have been mathematically proven - like EPR and Bells' inequality. A lot of people are unsatisfied with QM and looking for something more tangible - including physicists. Apparently pilot wave theory was never soundly rejected, the crowd just went in a different direction.
tjradcliffe 12y ago
The epistemic view of the wavefunction explains nothing, because it attempts to put all the quantum weirdness into "our knowledge of reality" rather than "reality". The whole argument seems to me silly, because the epistemic view requires that indistinguishable particles in fact are merely undistinguished particles. That is, the epistemic view requires that reality behaves quantum mechanically when no one is looking at it, whereas what we observe is that reality behaves quantum mechanically when no one can look at it. These are completely different situations. In one case it is what we do know that matters--which is an epistemic state--in the other it is what we can know that matters, which is an ontic state. And it turns out that reality does limit what we can know, and this has dramatic consequences. The clearest example of this is the heat capacity of solids. Heat capacity is a measure of how the temperature increases as you add heat energy to an object. The temperature is just the average energy per state in the material, so the heat capacity is a very direct measure of the number of states the collection of particles constituting a material can be in. It is very simple to show that particles that cannot be distinguished have a different number of states from particles that can be distinguished even if they don't happen to be at the moment. For example, consider two coins: if they can be distinguished they have four possible states, HH, HT, TH and TT. If they can't be distinguished then HT and TH are the same state (because there is no way, even in principle, even if you're god) to distinguish the left coin from the right coin. So there are only three states. You can do something similar with particles in a crystal lattice and show that they are indistinguishable. So it isn't a question of what we do know, but what we can know, and that is determined by ontology, not epistemology. The problem is that the ontology of the wavefunction is non-classical in the extreme. Because it is nonlocal it violates the law of non-contradiction from the perspective of locally causal beings (us). That makes it hard to understand, but it doesn't make it any less real. It just means reality is weird, which at this point in history shouldn't come as a huge surprise to anyone.
- FiatLuxDave 12y ago
  If you have not read Jayne's treatment of the Gibbs paradox, you may enjoy it. He addresses exactly the question of what it means to say we cannot distinguish particles from each other, and how 'cannot' really means 'cannot with our current state of knowledge and tools (including known fields, forces, etc).' See his analysis of two previously indistinguishable versions of Argon which become distinguishable through new tools. I realize that this is in a discussion of classical gas mixing physics, but the principle behind it is the same. So, in short, epistimology does matter if our level of knowledge does not exhaust the ontology. Link: http://bayes.wustl.edu/etj/articles/gibbs.paradox.pdf http://bayes.wustl.edu/etj/articles/gibbs.paradox.pdf

36 points by MichaelAO 12y ago

cma 12y ago

Full preprint article: http://arxiv.org/abs/1412.6213 http://arxiv.org/abs/1412.6213

jriordan 12y ago
Okay, I'm like 5 years old, but could you explain to me how this is available here for free, yet it costs $32 to read the final published version? I mean... this is fundamental human knowledge we are talking about... isn't there some better way of doing this than $32-a-pop?
- chriskanan 12y ago
  I have institutional access so I can't easily check, but I thought most or all Nature journals were now freely available: http://www.nature.com/news/nature-promotes-read-only-sharing-by-subscribers-1.16460 http://www.nature.com/news/nature-promotes-read-only-sharing...
  - jsprogrammer 12y ago
    From the link you posted: >All research papers from Nature will be made free to read in a proprietary screen-view format that can be annotated but not copied, printed or downloaded... >Annette Thomas, chief executive of Macmillan Science and Education, says that under the policy, subscribers can share any paper they have access to through a link to a read-only version of the paper’s PDF that can be viewed through a web browser. The person who linked the original article didn't use the special link to the "free, but proprietary" format.
- jsprogrammer 12y ago
  > but could you explain to me how this is available here for free, yet it costs $32 to read the final published version? Nature put a paywall on top of the article and then someone published the paywalled link on HN.
  - TTPrograms 12y ago
    I don't think that's quite right. Arxiv is a third party for academics to publish paper to get informal peer review outside of journals. My guess is that any edits made based on feedback from Nature's author network as well as the actual paper in the Nature format are under Nature's copyright. But yes, my understanding is that anybody with institution access to Nature can now share a link to the Nature publication.
    - jsprogrammer 12y ago
      Arxiv shows that the most recent revision of their document was published on January 20, 2015 (less than three weeks ago). I suppose it is possible that Nature's version is substantially different, but if that were the case it would cast a huge shadow of doubt over Nature's editorial process, as well as the claims made in the paper itself.
      - TTPrograms 12y ago
        The differences are likely a few small sentence additions, citations and comparisons to other work. I think the results would remain unaltered.
      - montecarl 12y ago
        Peer review and getting an article published takes at least a month if not half a year or longer. The preprint on Arxiv is most likely the final version of the paper they submitted for publication post peer review.
    - privong 12y ago
      > My guess is that any edits made based on feedback from Nature's author network as well as the actual paper in the Nature format are under Nature's copyright. Half correct. The authors retain copyright for changes due to feedback (peer review), as the authors themselves would have modified the paper to address any feedback. You are correct that the actual paper in the Nature formatting (and possibly associated copy-editing) are under Nature's copyright.

TTPrograms 12y ago

Every now and then people (on HN and elsewhere) like to rediscover pilot-wave theory and other similar ideas that posit that billiard-ball-like-particles still describe the universe. In these interpretations the wave-function, which is the probabilistic field that modern physics uses to model particles, is just an effective representation about our knowledge about the system. I tend to see these theories as paying a price of drastic increase in complexity to retain classical intuition about the system. The mainstream view in modern physics (for the past 30 or so years, I believe) is rather that the wavefunction is the fundamental description of reality, at least in the sense that classical effective knowledge theories are not plausible models. My interpretation of this paper on quick glance is that it is an attempt to experimentally demonstrate that only effective knowledge theories with very strange properties would suffice to explain observations, thus strengthening the "wave-function is fundamental" viewpoint.

jriordan 12y ago
What do you think is the best way to learn of these things, i.e. physics fundamentals and all in the modern state of knowledge? I ask from the perspective of a B.S. Physics cum computer software person... which, I think, might not be too far removed from the perspectives of most HN peeps...
- TTPrograms 12y ago
  Understanding current thought in the field for me is based on skimming many academic papers. That's kind of different, though. Getting a solid understanding of the fundamentals of the various fields is more of a coursework sort of thing. There's a bunch of quantum lectures on MIT's open courseware - anything Allen Adams will be great. If you're a book kind of person I believe I enjoyed Shankar but it's been a while.
phkahler 12y ago
>> Every now and then people (on HN and elsewhere) like to rediscover pilot-wave theory and other similar ideas that posit that billiard-ball-like-particles still describe the universe. Have you seen the experiments of Yves Couder? Hypothetically, if a model like that produced an exact mathematical equivalent of quantum theory wouldn't that be more satisfying? It would also destroy some of the notions that have been mathematically proven - like EPR and Bells' inequality. A lot of people are unsatisfied with QM and looking for something more tangible - including physicists. Apparently pilot wave theory was never soundly rejected, the crowd just went in a different direction.

tjradcliffe 12y ago

The epistemic view of the wavefunction explains nothing, because it attempts to put all the quantum weirdness into "our knowledge of reality" rather than "reality". The whole argument seems to me silly, because the epistemic view requires that indistinguishable particles in fact are merely undistinguished particles. That is, the epistemic view requires that reality behaves quantum mechanically when no one is looking at it, whereas what we observe is that reality behaves quantum mechanically when no one can look at it. These are completely different situations. In one case it is what we do know that matters--which is an epistemic state--in the other it is what we can know that matters, which is an ontic state. And it turns out that reality does limit what we can know, and this has dramatic consequences. The clearest example of this is the heat capacity of solids. Heat capacity is a measure of how the temperature increases as you add heat energy to an object. The temperature is just the average energy per state in the material, so the heat capacity is a very direct measure of the number of states the collection of particles constituting a material can be in. It is very simple to show that particles that cannot be distinguished have a different number of states from particles that can be distinguished even if they don't happen to be at the moment. For example, consider two coins: if they can be distinguished they have four possible states, HH, HT, TH and TT. If they can't be distinguished then HT and TH are the same state (because there is no way, even in principle, even if you're god) to distinguish the left coin from the right coin. So there are only three states. You can do something similar with particles in a crystal lattice and show that they are indistinguishable. So it isn't a question of what we do know, but what we can know, and that is determined by ontology, not epistemology. The problem is that the ontology of the wavefunction is non-classical in the extreme. Because it is nonlocal it violates the law of non-contradiction from the perspective of locally causal beings (us). That makes it hard to understand, but it doesn't make it any less real. It just means reality is weird, which at this point in history shouldn't come as a huge surprise to anyone.

FiatLuxDave 12y ago
If you have not read Jayne's treatment of the Gibbs paradox, you may enjoy it. He addresses exactly the question of what it means to say we cannot distinguish particles from each other, and how 'cannot' really means 'cannot with our current state of knowledge and tools (including known fields, forces, etc).' See his analysis of two previously indistinguishable versions of Argon which become distinguishable through new tools. I realize that this is in a discussion of classical gas mixing physics, but the principle behind it is the same. So, in short, epistimology does matter if our level of knowledge does not exhaust the ontology. Link: http://bayes.wustl.edu/etj/articles/gibbs.paradox.pdf http://bayes.wustl.edu/etj/articles/gibbs.paradox.pdf