Physicists Discover a Way Around Heisenberg's Uncertainty Principle

Hugh Pickens writes writes "Science Daily Headlines reports that researchers have applied a recently developed technique to directly measure the polarization states of light overcoming some important challenges of Heisenberg's famous Uncertainty Principle and demonstrating that it is possible to measure key related variables, known as 'conjugate' variables, of a quantum particle or state directly. Such direct measurements of the wave-function had long seemed impossible because of a key tenet of the uncertainty principle — the idea that certain properties of a quantum system could be known only poorly if certain other related properties were known with precision. 'The reason it wasn't thought possible to measure two conjugate variables directly was because measuring one would destroy the wave-function before the other one could be measured,' says co-author Jonathan Leach. The direct measurement technique employs a 'trick' to measure the first property in such a way that the system is not disturbed significantly and information about the second property can still be obtained. This careful measurement relies on the 'weak measurement' of the first property followed by a 'strong measurement' of the second property. First described 25 years ago, weak measurement requires that the coupling between the system and what is used to measure it be, as its name suggests, 'weak,' which means that the system is barely disturbed in the measurement process. The downside of this type of measurement is that a single measurement only provides a small amount of information, and to get an accurate readout, the process has to be repeated multiple times and the average taken. Researchers passed polarized light through two crystals of differing thicknesses: the first, a very thin crystal that 'weakly' measures the horizontal and vertical polarization state; the second, a much thicker crystal that 'strongly' measures the diagonal and anti-diagonal polarization state. As the first measurement was performed weakly, the system is not significantly disturbed, and therefore, information gained from the second measurement was still valid. This process is repeated several times to build up accurate statistics. Putting all of this together gives a full, direct characterization of the polarization states of the light."

Re:Does this break Quantum Key Distribution?

[johndoe42]

No, because the summary is (as usual) thoroughly overstated. This experiment, like any other form of quantum state tomography [wikipedia.org] lets you take a lot of identical quantum systems and characterize them. For it to work, you need a source of identical quantum states.

As a really simple example, take a polarized light source and a polarizer (e.g. a good pair of sunglasses). Rotate the polarizer and you can easily figure out which way the light is polarized. This is neither surprising nor a big deal -- there are lots of identically polarized photons, so the usual uncertainty constraints don't apply.

The whole point of QKD (the BB84 and similar protocols) is that you send exactly one photon with the relevant state. One copy = no tomography.

This is not a way *around* Heisenberg

[Wrath0fb0b]

What they are doing is assuming that their light source is broadly uniform and averaging over the double-measurement (which is clever, no doubt). So we still haven't learned anything about a particular photon that violates the uncertainty principle, only something about the entire population. If we assume that the population is uniformly polarized (which is reasonable in this case) then we can conclude that the average reflects the properties of the individual photons. If the population was not uniform, however, then the average tells us very little about the properties of the individual photons.

And before someone too clever tries to argue that you can take a single input photon and make multiple copes and send them through this process to get results about that one photon, there is the No Clone Theorem [wikipedia.org] to here to prevent that maneuver.

So really they haven't gone around Heisenberg (which talked only about individual wave-functions) but used multiple compound measurements and an assumption about the properties of the group to infer something that Heisinberg says they can't measure directly -- which is quite clever but Herr Doctor's principle still stands quite strong.

Not a violation of the uncertainty principle

[mpoulton]

Like many non-rigorous descriptions, the summary makes the mistake of describing the uncertainty principle as if it is a measurement problem, where the lack of precision somehow arises from inadequate measurement technology. This is not a correct statement of the uncertainty principle. The fundamental issue is that the conjugate variable values are linked on a quantum level, such that there is a certain amount of natural, inherent uncertainty in their collective values due to the statistical/wavelike nature of the quantum particle. With perfect measurement, there is still uncertainty in the pair of values for any conjugate variables because the uncertainty lies in the actual values themselves. Position and momentum are the quintessential conjugate pair. The Heisenberg uncertainty principle is sometimes framed as the idea that you cannot know the speed and position of a particle at the same time. But it's more correct to say that a particle does not HAVE an exact speed and position at the same time. This weak measurement technique is certainly useful and interesting since it allows some observations of wavefunctions without collapse, but it does not actually allow the measurement of conjugate variables more precisely than the uncertainty principle allows - because the values themselves do not exist more precisely than that.

*This description is based one one of the multiple interpretations of quantum mechanics, and probably does not accurately represent physical reality, only our human understanding of a part of reality that we have not really figured out completely yet.

Re:Not a violation of the uncertainty principle

[pclminion]

For those with a signal processing background, it can be explained like this. The conjugate pair of momentum and position are related to each other by the Fourier transform -- the Fourier transform of the wavefunction in spatial coordinates yields the wavefunction in momentum coordinates. Anybody who has worked with a Fourier transform knows that if the input is band-limited, the output will not be, and vice versa. To know the position of a particle with exactness implies that its wavefunction is impulse-like in the spatial domain, which causes the momentum wavefunction to be a wave that extends infinitely throughout momentum-space. When you squeeze the bandwidth in one domain it grows in the other. Because the Fourier transform of a Gaussian is another Gaussian, a particle with Gaussian distribution in either space or momentum-space constitutes the most localizable wavefunction one could possibly achieve. The limit of the resolution is given by the Heisenberg relation, but this is a purely mathematical result, having nothing to do with measurement technique.

Re:Br Ba

[pjt33]

No, Heisenberg bounds the product of the errors in the measurements of the two by means of a Schwartz inequality: i.e. if you measure one very precisely, you will get a big error in your measurement of the other one.

Re:It's the fault of the stupid haedline

[sonnejw0]

They are repeating the measurement multiple times on a stream of photons. They're not measuring the same particle repeatedly, they're not even close to overcoming the uncertainty principle.