Originally Posted by

**DParlevliet**
I know, but still: the earlier argument of detecting in detector 2 does not fit anymore... and other detailed explanation.

But going back to the original argument, what is the exact definition of "knowing-which-path"? The general principle is that when you know the which-path (particle property) there cannot interference (wave property). But look to a simple one-slit measurement:

You know which path, but still the photon can be detected sideways from the slit because of diffraction. And diffraction is also a wave property, also based on interference (Huygens principle, calculated on waves travelling in a medium). So do we have here a knowing-which-path together with wave property?

But back to the first question. We know which-path in the macro world, but in the quantum world the slit is very wide. There are numeroes paths the small photon can go though the slit, and we don't know which. So perhaps still a situation of not knowing the path?

So what is the exact defination of "knowing the path"?

Personally, I regard the "wave-particle duality" as an oversimplification. The way I see it, it's never "particles", just waveforms of various shapes, including very narrow pulses, which could be considered as "particles". So for the single-slit diffraction, it's not a particle, but a waveform that is as wide as the slit, and this produces a single-slit interference pattern. For the double-slit experiment, the waveform is a pair of slit-shaped pulses, and measuring which slit the photon passed through means eliminating one of the slit-shape pulses.

The crucial aspect of quantum mechanics is that when one performs a measurement,

*any* measurement, of all the possible classically distinct results (with probabilities given by the wavefunction), only

*one* is realised. This is the true meaning of a "particle" in quantum mechanics. Thus, the true nature of a particle depends entirely on the property that is actually being measured, noting that it is in the classical realm that the different possible results distinguish themselves. That is, a measuring device is a macroscopic object and it is the possible distinct states of that macroscopic object that constitute the possible measured results, and it is this that represents the measured quantum state.

As for what it means to "know the path", it is important to note that it is not actually the

*knowledge* of the path that destroys the interference. Strictly speaking, it is orthogonalising the quantum states corresponding to each of the paths that destroys the interference. Typically, this means making each of the paths classically distinguishable, even if this is not the result of an explicit measurement. This means that destroying interference is rather easy, and maintaining interference is rather difficult.