Albert Einstein played a key role in launching quantum mechanics through his theory of the photoelectric effect but remained deeply bothered by its philosophical implications. And though most of us still remember him for deriving E=MC^2, his last great contribution to physics was actually a 1935 paper, coauthored with his young colleagues Boris Podolsky and Nathan Rosen. Regarded as an odd philosophical footnote well into the 1980s, this EPR paper has recently become central to a new understanding of quantum physics, with its description of a strange phenomenon now known as entangled states.
The paper begins by considering a source that spits out pairs of particles, each with two measurable properties. Each of these measurements has two possible results of equal probability. Let's say zero or one for the first property, and A or B for the second. Once a measurement is performed, subsequent measurements of the same property in the same particle will yield the same result. The strange implication of this scenario is not only that the state of a single particle is indeterminate until it's measured, but that the measurement then determines the state.
What's more, the measurements affect each other. If you measure a particle as being in state 1, and follow it up with the second type of measurement, you'll have a 50% chance of getting either A or B, but if you then repeat the first measurement, you'll have a a 50% chance of getting zero even though the particle had already been measured at one. So switching the property being measured scrambles the original result, allowing for a new, random value.
Things get even stranger when you look at both particles. Each of the particles will produce random results, but if you compare the two, you will find that they are always perfectly correlated. For example, if both particles are measured at zero, the relationship will always hold. The states of the two are entangled. Measuring one will tell you the other with absolute certainty.
But this entanglement seems to defy Einstein's famous theory of relativity because there is nothing to limit the distance between particles. If you measure one in New York at noon, and the other in San Francisco a nanosecond later, they still give exactly the same result. But if the measurement does determine the value, then this would require one particle sending some sort of signal to the other at 13,000,000 times the speed of light, which according to relativity, is impossible. For this reason, Einstein dismissed entanglement as "spuckafte ferwirklung," or spooky action at a distance. He decided that quantum mechanics must be incomplete, a mere approximation of a deeper reality in which both particles have predetermined states that are hidden from us. Supporters of orthodox quantum theory lead by Niels Bohr maintained that quantum states really are fundamentally indeterminate, and entanglement allows the state of one particle to depend on that of its distant partner.
For 30 years, physics remained at an impasse, until John Bell figured out that the key to testing the EPR argument was to look at cases involving different measurements on the two particles. The local hidden variable theories favored by Einstein, Podolsky and Rosen, strictly limited how often you could get results like 1A or B0 because the outcomes would have to be defined in advance. Bell showed that the purely quantum approach, where the state is truly indeterminate until measured, has different limits and predicts mixed measurement results that are impossible in the predetermined scenario. Once Bell had worked out how to test the EPR argument, physicists went out and did it.
Beginning with John Clauster in the 70s and Alain Aspect in the early 80s, dozens of experiments have tested the EPR prediction, and all have found the same thing: quantum mechanics is correct. The correlations between the indeterminate states of entangled particles are real and cannot be explained by any deeper variable. The EPR paper turned out to be wrong but brilliantly so. By leading physicists to think deeply about the foundations of quantum physics, it led to further elaboration of the theory and helped launch research into subjects like quantum information, now a thriving field with the potential to develop computers of unparalleled power.
Unfortunately, the randomness of the measured results prevents science fiction scenarios, like using entangled particles to send messages faster than light. So relativity is safe, for now. But the quantum universe is far stranger than Einstein wanted to believe.
