From: "Jesse Mazer" > Would this experimental result actually be predicted by the quantum > formalism, though? It sounds like they had a setup similar to the > double-slit experiment and found a small amount of interference even when > they measured which hole the particle traveled through, but I thought the > quantum formalism predicts that interference would be completely destroyed > by such a measurement.

There is a lot of confusion about all that. I hope I do not make more damages here! There are also many different versions of Bohr's complementarity principle. Complementarity of what? Waves (there are no waves in matrix mechanics!) and particles? Interference pattern and "which way"? Continuous and discontinuous? Localization and superposition? Separability and unitarity? Reversibility and irreversibility? The modern view says that ... "The superposition of amplitudes is only valid if there is no way to know, even in principle, which path the particle took. It is important to realize that this does not imply that an observer actually takes note of what happens. It is sufficient to destroy the interference pattern, if the path information is accessible in principle from the experiment or even if it is dispersed in the environment and beyond any technical possibility to be recovered, but in principle 'still out there'". Anton Zeilinger, Rev. Mod. Phys., 1999, page S-288 "In an experiment the state reflects not what is actually known about the system, but rather what is knowable, in principle, with the help of auxiliary measurements that do not disturb the original experiment. By focusing on what is knowable in principle, and treating what is known as largely irrelevant, one completely avoids the anthropomorphism and any reference to consciousness that some physicists have tried to inject into quantum mechanics" Leonard Mandel, Rev. Mod. Phys., 1999, p. S-274. So, the key word now is "indistinguishability". Must this "indistinguishability" be absolute? What does it happen in case of partial "indistinguishability"? (Anticipated answer: there is a smooth transition between particle-like and wave-like behaviour). In 1979, Wootters and Zurek (Complementarity in the double-slit experiment: Quantum nonseparability and a quantitative statement of Bohr's principle, PR, D-19, 1979, p. 473-484) presented a famous gedanken experiment, showing that photons still have a wave-like behaviour even if their paths are predicted almost (say: 99%) certainly. The set-up, in the gedanken, was essentially a single-slit plus a double-slit; and also a double-slit plus a specific 'textured' screen capable of detect and record both the interference pattern and the 'which way'. Yes this is possible. Coupling Wheeler's 'delayed choice' and the above gedanken experiment, Wim Rietdijk wrote (circa 1982) an interesting paper. Very shortly, QM explains the two-slit interference via Heisenberg principle. Hence the slits measure the position of the 'object'; because of this measurement there is a scattering; |<p(y)|psi|^2 gives the probability function for the 'object' emerging from the slits with momentum p(y); this probability function causes the interference pattern. Thus - that is important - after the 'object' has passed through the two-slit, the probability function |<p(y)|psi|^2 is fixed. And - second important point - there is a principle of conservation of momentum. Thus, nothing can change that fixed momentum (rectius: that fixed probability function). Now comes the weirdness. After the 'object' has passed the two-slit, we have *still* some time to choose if we wish to detect the 'welcher weg' (wich way, which path) the 'object' took, or if we wish to record just the 'impact' of the 'object' on the screen or, in general, if we wish to get both, the 'welcher weg' and the 'interference pattern' at the same time (this is technically possible, provided we use a screen with a special 'texture'). Here is the weirdness: does QM say that any knowledge of the 'welcher weg' causes the loss of the interference pattern? Yes? Does Feynman say this in his Lectures? Ok. Thus QM says that the the probability function |<p(y)|psi|^2, already fixed at the two-slit level, is a function of our later, delayed, free choice of a specific detector (of the interference pattern only; of the interference pattern and the 'wich path' at the same time). Coming back to the point of that "absolute" indistinuishability. Greenberger and Yasin wrote down the relation, P^2 + V^2 = 1, where P is the probability for the electron (or photon) taking one of the two possible paths, and V the visibility of the fringes (interference pattern). http://arxiv.org/abs/quant-ph/9908072 http://arxiv.org/abs/quant-ph/0311179 http://arxiv.org/abs/quant-ph/0201026 In other words, the Greenberger and Yasin relation states that the "entity" (electron, photon, etc.) has a double nature (wave-like, particle-like) and there is a "smooth" transition between one and the other nature. So, the "indistinuishability" is not an absolute, by experiment. See this specific new test http://www.arxiv.org/abs/quant-ph?0404013 Of course there are interesting perspectives using "weak" measurements http://www.arxiv.org/abs/quant-ph?0310081 and photons of different energy (wave-lenght) http://www.arxiv.org/abs/quant-ph/0304086 and so on (non-local two-slit interferometers, interference between two correlated sources, interference between two uncorrelated sources, interference in time of any kind, quantum beats, etc.). In summa. There is no new at all. After all Bohr even incorporated Chinese traditional Yin-Yang Symbol, and related smooth transition, into his Coat of Arms. http://www.upscale.utoronto.ca/GeneralInterest/Harrison/Complementarity/Comp Copen.html