When Isaac Newton published his theory of universal gravitation in 1686, he knew he'd have to confront a few critics. Like a ghost stretching its arms across empty space, Newton's theory described the gravitational attraction between two masses, say, the Sun and the Earth, as a mysterious force that acted instantaneously between them.

How could the Sun influence the Earth, and the Earth the Sun, without direct contact?

This is the challenge of instantaneous "action-at-a-distance," unexplained in Newton's theory. Newton tried to preempt his enemies' attacks by adding a disclaimer to his work. He argued that his theory was so good at explaining so many things that to ask where the force of gravity comes from, or how it acts across space was, irrelevant. Rather, he would "feign no hypotheses" as these could not be part of "experimental philosophy."

In 1915 Albert Einstein moved things forward with a radical new way of thinking about gravity. According to his general theory of relativity, what we call a gravitational force is due to the bending of space around a mass. Like a stretchy fabric, space gets deformed to a larger or smaller extent depending on the mass. Planets orbit the Sun because the Sun's huge mass bends space in such a way that the shortest path around it is in the shape of an ellipse.

Einstein brilliantly exorcised the ghost from Newton's action-at-a-distance, showing that in effect gravity is a perfectly local force. Furthermore, gravitational effects don't act instantaneously through space but propagate with the speed of light. If the Sun disappeared right now you'd only find out in 8.3 minutes, the time it takes for light (and gravitational disturbances) to propagate from there to here.

Unfortunately, before Einstein could kick back and relax, quantum physics brought back the action-at-a-distance ghost.

It all started when Erwin Schrödinger proposed his wave equation in 1926, originally to describe how electrons orbit the atomic nucleus without falling in. It was clear that the electron couldn't be described solely as a particle; de Broglie had shown that it could actually also be a wave, giving rise to the so-called wave-particle duality or, as Bohr would state it, two complementary descriptions of physical reality.

Soon it was clear that Schrödinger's wave was not an electron, or anything material, but a probabilistic description of what the electron could do. As I wrote here last week, the fundamental equation of quantum physics does not describe things. The probabilistic wave represented potentialities, possible results of a measurement with a certain probability: when measured, the electron could be here or there, with this or that probability.

Strictly speaking, the very act of measuring the electron's position would create it in a certain spot in space. Mathematically, the act of measurement means that the wave describing the potentialities collapses into a single point, the place where the electron is found. To Einstein's great irritation, the wave would collapse instantaneously! And how would distant parts of the wave "know" when to collapse? Was there a signal travelling faster than light? Was action-at-a-distance back?

Things got worse in 1935 when Einstein, Podolsky and Rosen tried to show that the quantum description of Nature was incomplete, that a better theory was needed. Let me summarize with an adapted version of their challenge, known as the EPR paradox.

When light is polarized its associated wave goes up and down in the same direction, as when we ride a horse. Photons of polarized light share this polarization.

Let's imagine that, in an experiment, a source of light created a pair of polarized photons traveling in opposite directions, say east and west. Imagine that two physicists, Alice and Bob, each stood with a detector at 10 yards from the source. Alice at the east and Bob at the west. Since photons travel at the speed of light, the pair would detect photons arriving at their detectors at the same time.

Now imagine that each of the detectors could detect two possible polarizations, vertical or horizontal. The light source is such that when the pair of photons leaves it, they have the same polarization, either vertical or horizontal. Alice and Bob don't know which until they measure it. Let's say Alice measures vertical; Bob will measure vertical too. If Alice measures horizontal, so will Bob. There will be 50 percent probability that the photon will be found in either state of polarization. So far so good.

Alice decides to move a bit closer to the source. She measures a photon with vertical polarization. Immediately, she knows that Bob's photon will also have vertical polarization, *before the photon even gets to Bob's detector*. But according to quantum mechanics, you can only tell the state of something by looking. And since nothing can travel faster than the speed of light, Alice apparently influenced Bob's photon instantaneously without interacting with it.

She could have been even more cunning and inverted the polarization of her photon before Bob's got to him. Again, she would know that Bob's photon would have the same polarization as hers. Einstein called this "spooky action-at-a-distance," the mysterious and wondrous quantum ghost. Given what he had done to Newton's ghost, we can see why he was so keen on getting rid of this one too.

Nowadays such entangled pairs of particles are created and analyzed in laboratories around the world. One of my favorites is Anton Zeilinger's lab at the University of Vienna. The photons in the experiment above form an entangled pair and, thus, a single unit. Changing one instantaneously influences the other, irrespective of how far the two are from each other. This month, Zeilinger's group, in collaboration with the U.S. National Institute of Standards and Technology, performed a state-of-the-art experiment confirming that Nature does work through spooky actions-at-a-distance.

What's really disconcerting is that Einstein's hope for a local theory explaining what's going on — exorcizing the quantum ghost — has been shot down. The experiment rules out local theories of quantum mechanics to explain instantaneous action-at-a-distance. What physicists call "nonlocality," influences from elsewhere acting immediately on separated entangled pairs, is a ghost that seems to be real.

Reality is not just strange. It's far stranger than we can suppose.

*You can keep up with more of what Marcelo is thinking on Facebook and Twitter: @mgleiser*