# Does Feynman's approach to QED render the observer problem moot?

December 5, 2005 10:05 AM Subscribe

This site:
http://www.skepticreport.com/print/quantum-p.htm
...suggests that Feynman's 'sum over all histories' approach to Quantum Electro-Dynamics has sidestepped the well known 'observer problem' in quantum mechanics (exemplified by the Wigner's Friend paradox). Is this true?

According to some conventional approaches to quantum mechanics, a waveform collapses when a particle is 'observed'. But the meaning of 'observed' is moot. According to the Copenhagen Interpretation is means something along the lines of 'interacts with a classical (as opposed to quantum) system'. But since all systems can be regarded as quantum in some sense, it doesn't really answer the question. In the Schroedinger's cat paradox, a cat whose survival or death depends on the decay of an atom is in a strange alive/dead superimposition-state until an observer opens a box. If you add a human being watching the cat to the room, and close the door ('Wigner's Friend') then the human being is likewise in a superimposition of states ('believing the cat is dead'/'believing the cat is alive') until the observer opens the door and observers him/her. But of course you can keep adding doors forever, until you finally require some ur-observer (God?) to observer the entire universe and force the waveforms to collapse.

Other explanations include the many-universes theory and the idea that consciousness itself somehow defines an 'observer'.

The Skeptic site claims that Feynman's approach of summing over all possible histories to arrrive at the probability of a particular event sidesteps the observer problem, but my understanding is that all it does is provide a neat way of calculating the probabilities to an arbitrary accuracy. Can somone help me?

According to some conventional approaches to quantum mechanics, a waveform collapses when a particle is 'observed'. But the meaning of 'observed' is moot. According to the Copenhagen Interpretation is means something along the lines of 'interacts with a classical (as opposed to quantum) system'. But since all systems can be regarded as quantum in some sense, it doesn't really answer the question. In the Schroedinger's cat paradox, a cat whose survival or death depends on the decay of an atom is in a strange alive/dead superimposition-state until an observer opens a box. If you add a human being watching the cat to the room, and close the door ('Wigner's Friend') then the human being is likewise in a superimposition of states ('believing the cat is dead'/'believing the cat is alive') until the observer opens the door and observers him/her. But of course you can keep adding doors forever, until you finally require some ur-observer (God?) to observer the entire universe and force the waveforms to collapse.

Other explanations include the many-universes theory and the idea that consciousness itself somehow defines an 'observer'.

The Skeptic site claims that Feynman's approach of summing over all possible histories to arrrive at the probability of a particular event sidesteps the observer problem, but my understanding is that all it does is provide a neat way of calculating the probabilities to an arbitrary accuracy. Can somone help me?

I don't think the sum over histories approach sidesteps the observer problem. It does sidestep wave-particle duality though, which is related. In some interpretations of QM, the "particle" is actually a wave (wave-function) which collapses to a point-particle after observation. In Fenyman's approach, the particle is always a particle, but it travels over every conceivable path and the totality of the paths make it interfere with itself like a wave. Disclaimer: I am not a quantum mechanic.

posted by zanni at 11:41 AM on December 5, 2005

posted by zanni at 11:41 AM on December 5, 2005

Have you been following all of these, and this. The issue has been addressed directly and indirectly there. Someone brought up Bohmian mechanics, which is one approach (that I looked into), I have no idea how many other approaches there are - many, though.

posted by Chuckles at 11:51 AM on December 5, 2005

posted by Chuckles at 11:51 AM on December 5, 2005

zanni is essentially right. The Feynman approach already assumes a "destination" in point B but doesnt have to explain the "collapse" at point B since the particle was always a particle in a sense. The same paradoxes are still there, they've just been reformulated (why is a particle evaluating all possible paths? And why to that destination B?) It is a beautiful approach though, especially since it basically cuts Time out of the picture.

posted by vacapinta at 11:55 AM on December 5, 2005

posted by vacapinta at 11:55 AM on December 5, 2005

please keep in mind that the cat's alive-or-dead state is

posted by delmoi at 12:16 PM on December 5, 2005

**A METAPHORE**. tia.posted by delmoi at 12:16 PM on December 5, 2005

I don't see how any of this can overcome the Heisenberg uncertainty principle. You're still shooting photons (or some other method) into something to observe it, thus changing its state somewhat.

posted by skallas at 12:25 PM on December 5, 2005

posted by skallas at 12:25 PM on December 5, 2005

*please keep in mind that the cat's alive-or-dead state is A METAPHORE.*

what is this supposed to mean? could you elaborate?

posted by sergeant sandwich at 12:26 PM on December 5, 2005

*please keep in mind that the cat's alive-or-dead state is A METAPHORE.*

what is this supposed to mean? could you elaborate?

what is this supposed to mean? could you elaborate?

Well, he means metaphor. But the cat example was how Schrodinger described what's weird about quantum particles. Under no circumstances is it supposed to describe the behaviour of a cat in a box with a vial of poison gas. In other words, don't try this at home.

posted by lumpenprole at 12:39 PM on December 5, 2005

Uncertainty does not come from disturbing what you are measuring (although that does happen). Uncertainty is because the thing you are measuring either has definite velocity or definite position. It cannot have both. At least that's what Feynman told me.

posted by jockc at 3:29 PM on December 5, 2005

posted by jockc at 3:29 PM on December 5, 2005

This thread is closed to new comments.

justreading about this last night.see if you can find a copy of goswami's

quantum mechanicsand have a look at the last chapter, called something like 'quantum paradoxes: the unfinished chapter'. he spends a good deal of time talking about precisely this issue.i think the general conclusion is that there does not exist (yet) a resolution to this paradox of quantum mechanics. goswami's speculation is that consciousness itself might be a quantum mechanical phenomenon, or that the behavior of the observer's mind may ultimately be found to exhibit behavior not yet seen in other quantum mechanical systems.

nevertheless it's a good read!

posted by sergeant sandwich at 10:45 AM on December 5, 2005