Monday, November 28, 2011

Feynman on QM in 1964

Whenever I want to mention Richard Feynman's attitude to the foundations of quantum mechanics, I typically point to this four-minute interview with an older Feynman. He says that physics at the fundamental quantum level is so fantastically different from anything we have seen before.

There's still a school of thought that doesn't want to believe that those rules are fundamental and wants to find out some mundane things beneath all the phenomena. This effort is due to their deep prejudice, Feynman says, and they will be defeated because Nature's imagination is much greater than Man's. She will never let us relax.

Decades later, those prejudiced people still haven't made any insight about Nature that would go in their preconceived direction but as the human population has dumbed down, it's also true that they have failed to be defeated. This anti-quantum stupidity is at least as alive as it was during Feynman's life.

An 8-minute beginning of the lecture

However, it's also interesting to listen to pretty much the same things that Feynman said in his 1964 Messenger Lectures at Cornell. The degree of constancy of his opinions is remarkable – and one could say that when it comes to topics that rapidly evolved during his later years, this constancy was too much of a good thing.

If you want to see the whole lecture, open your Microsoft Internet Explorer (with Silverlight) and go to Bill Gates' Project Tuva in it. Or open the Tuva page in Chrome and click at the icon of the IE Multi Tab extension that you previously installed. Then click at Feynman's (not Gates') photograph and choose lecture 6, Probability and Uncertainty, the Quantum Mechanical View of Nature.

In the beginning, he says that it shouldn't have been unexpected that the fundamental laws of Nature look ever more unintuitive. But the students in the room shouldn't try to "understand" what Feynman is saying in the sense of trying to create a model in terms of something they already know. There's nothing like that. They should appreciate Nature how She really is. And She's delightful and sexy and whoever doesn't see this sexiness is a physics faggot.

The bulk of the lecture is dedicated to the double-slit experiment which knows about all the delightful aspects of the particles' quantum behavior. And indeed, it does.

Around 50:00, he also explains the concept of "hidden-variable" theories and demonstrates that this is not a possible description of Nature. The reason is simple: if it were in principle possible to predict in advance whether the electron would be observed in the slit #1 or slit #2 (assuming that the photon trackers are added), then it would inevitably imply that the experiment in the absence of photon trackers has to produce the \(N_1+N_2\) dull juxtaposition on the screen instead of the nice \(N_{12}\) interference pattern.

You don't need any idealized experiments with qubits; the double-slit experiment is enough to falsify all such fundamentally misguided theories. The probabilistic character of the predictions seems to be Nature's intrinsic property; it is not due to the lack of knowledge of the internal wheels and gears. As someone said, Nature Herself doesn't know which way it will go.

Deliciously enough, near 52:10, Feynman mentions that a (pompous) philosopher once said (in a deep authoritarian voice) that it is necessary for the processes in science to produce the same results from the same conditions. Well, they don't, Feynman responds. ;-)

He even constructs a thought experiment in which the World War III (something important) depends on this quantum random outcome. No amount of science can settle it, anyway. The future of the world is unpredictable. Science isn't about obeying arbitrary philosophical preconditions; it depends on the ability to do experiments, on honesty in reporting them and on intelligence in interpreting them. However, it must be a non-dogmatic kind of intelligence that is not sure ahead of time what the results should be. Finite bias is OK because a finite amount of accumulated evidence will ultimately push a biased person to give up and accept the truth. ;-)

The debunking of various influential pseudoscientific hypotheses such as the hidden-variable theories should become a standard part of the QM and other courses, I think. Throughout 47 years after Feynman's Messenger Lectures, the public has made no positive progress in understanding the basic framework of modern physics. Quite on the contrary.

Alain Aspect's colloquium at Technion

You may also listen to a newer talk:

Alain Aspect, the well-known boss of an Atom Optiques institute in Fʁɑ̃nce, gave a 55-minute talk at the Isʁael's Technion in August 2011. It focuses on the wave-paʁticle duality, pʁetty much the same topic as Feynman's lectuʁe above. Aspect reviewed the history of light as particles and waves. Then he discussed experiments with individual photons; Feynman was the only one who immediately told him what would happen in the experiments, Aspect recalls. Finally, we learn about Wheeler's delayed choice experiments. Bohr's complementarity principle is shown in action.


  1. "He says that physics at the fundamental quantum level is so fantastically different from anything we have seen before," yet he still admits we can understand it in analogy to things we have seen before. New knowledge has to be partly based upon what we already know.

  2. Nope, Alan, you didn't listen carefully. Once he mentioned "analogies" for the first time, he immediately corrected himself within seconds, saying that he will need to use "analogies and contrasts" because "analogies" couldn't work. And indeed, they don't work.

    Any analogy with anything in the real world inevitably predicts wrong results in the double slit experiment, as he discusses in detail.

    So the knowledge in QM may depend on what we knew before but one must add "NOT" to the previous knowledge because quantum mechanics is, well, the first physics theory in the history of science that isn't classical so its framework isn't analogous to anything in classical physics.