Trespassing on Einstein's Lawn (56 page)

On one hand, I wasn't surprised. Or at least I shouldn't have been. Everything I had learned up until now had fully prepared me for this moment. Is the elephant dead and crispy outside the horizon or alive and terrified inside? It depends who you ask. There's no God's-eye view that contains the “truth” of the matter. The “truth” of the matter is observer-dependent. On the other hand, Rovelli did seem to be taking the observer-dependent thing to a whole new level. He was making
everything
observer-dependent, and reinventing quantum mechanics in the process.

Fundamental physics proceeds by paradox. It always has. It was a paradox that led Einstein to relativity: the laws of physics had to be the same for everyone and, given the relational motion of light, the laws of physics couldn't be the same for everyone. A paradox led Polchinski to D-branes: open strings had to obey T-duality and, given their boundary conditions, open strings couldn't obey T-duality. Another paradox led Susskind to horizon complementarity: information had to escape a black hole and, given relativity, information couldn't escape a black hole. And yet another led the entire physics community to wonder whether each observer has his or her own quantum description of the world: entanglement had to be monogamous and, given the equivalence principle, entanglement couldn't be monogamous.

There's only one way to resolve a paradox—you have to abandon some basic assumption, the faulty one that created the paradox in the first place. For Einstein, it was absolute space and time. For Polchinski, it was the immovability of the submanifold to which the open strings attached. For Susskind, it was the invariance of spacetime locality. For everyone involved in the firewall mess, it was the idea that quantum entanglement is observer-independent.

Quantum mechanics short-circuits our neurons because it presents yet another paradox: cats have to be alive and dead at the same time, and, given our experience, cats can't be alive and dead at the same time. Rovelli resolved the paradox by spotting the inherently flawed assumption: that there is a single reality that all observers share. That you can talk about the world from more than one perspective simultaneously. That there's some invariant way the universe “really is.”

I called my father and we discussed Rovelli's paper for hours, debating its implications for the meaning of ultimate reality until the Sun came up. Or until the Sun came up
relative to me.

Rovelli's relational quantum mechanics, which rendered wavefunction collapse observer-dependent, allowed him to tackle the thought experiment that Einstein himself had waged against quantum mechanics, hoping it would rip the theory apart at the seams: EPR.

A staunch realist, Einstein resented quantum theory's claim that somehow a particle doesn't have properties until it's measured. If quantum mechanics is probabilistic, he said, the probabilities reflect our subjective ignorance, not some objective uncertainty in reality itself.

Einstein (E), along with Boris Podolsky (P) and Nathan Rosen (R), proposed a thought experiment to prove it. The gist of the EPR experiment was simple: you have two quantum-entangled particles, say an electron and a positron. Because they're entangled, the two particles are described by a single wavefunction, which might have zero total spin. Consequently, if the electron's spin is measured up, the positron's must be down, and vice versa. Their spins have to be anti-correlated so that they always add to zero.

An entangled electron-positron pair is created somewhere in the middle of Connecticut. The particles part ways. One arrives at my door here in Boston; the other makes its way down toward Philadelphia. I decide to determine the electron's spin. Spin can be measured along any spatial direction:
x, y
, or z. I choose
x:
my electron's spin is up. Meanwhile, in Philadelphia, my father is about to measure the positron's spin along the
x
-axis. But the outcome of his measurement is already determined: it has to be down. He makes the measurement a mere fraction of a second later. Sure enough, it's down.

Einstein had a serious problem with this. How did my father's particle “know” about my measurement across several state lines? Any signal that my particle could have sent to my father's would have had to travel faster than light to get to Philly in time for the measurement. But of all people, Einstein wasn't about to allow for superluminal
speeds, or what he called “spooky action-at-a-distance.” The only reasonable explanation, according to EPR, was that the particles each had well-determined spins all along: the electron's spin was up before I measured it and the positron's spin was down from the start. There was some way that things
really were
, independent of either of our observations. After all, we could have chosen to measure our spins along the
y
or
z
axis—so the particles had to have set spin values along each axis from the get-go. These determined outcomes, or “hidden variables,” aren't reflected in the formalism of quantum mechanics. Ergo, said EPR, quantum mechanics is incomplete. Probabilities represent uncertainty in our knowledge, and not in reality itself.

Unfortunately for Einstein, John Stewart Bell shattered EPR's reality. He calculated that any hidden-variables theory would produce the wrong probabilities for the outcomes of multiple measurements—unless, that is, the hidden variables operated via spooky action-at-a-distance. The only way to save a reality out there, independent of observers, was to violate the locality at the heart of relativity.

Bell's theorem was tested over and over in different ways in labs around the world, all of which came up with the same result: Einstein was wrong. After a particularly damning lab test in 2007,
Physics World
ran an article headlined “Quantum Physics Says Goodbye to Reality.”

Even though Einstein had set the whole quantum revolution in motion, he couldn't accept what the theory was telling him about reality. You'd think he would have learned to ignore his philosophical prejudices after the whole expanding-universe fiasco. But no. He wanted to retreat back behind that thick plate-glass window where he could passively observe a reality that couldn't care less that he was observing it. But it was too late. Quantum mechanics had already smashed the glass, and there was Bell, stomping on the shards.

Bell showed that my electron has no defined spin—it's in a superposition of spin up and spin down—until it randomly chooses a value upon measurement. By randomly choosing a value, it also sets the value for my father's positron.
Instantaneously.
Faster than light. Since there's no way anyone could use this superluminal effect to transmit information, it wasn't a flagrant violation of relativity. But it was walking a pretty fine line.

Ever since Bell, physicists have surrendered to spooky action-at-a-distance. That's just the weird way things are, they said. But it never seemed quite right. And now, reading another paper by Rovelli, I understood exactly why.

“Einstein's reasoning requires the existence of a hypothetical superobserver that can instantaneously measure the state of [Amanda] and [her father],” the paper stated. “It is the hypothetical existence of such a nonlocal superbeing, and not quantum mechanics, that violates locality.”

The point was this: the EPR problem arises because it seems like by collapsing the wavefunction of my electron, I mysteriously collapse the positron's wavefunction several states away. But relational quantum mechanics tells a different story. When I measure my electron, its wavefunction collapses
relative to me.
As far as my dad is concerned, the electron's wavefunction hasn't collapsed at all, and I'm in a superposition of “having measured spin up” and “having measured spin down.” Nothing superluminal is going on. He can take a trip up to Boston and collapse that superposition to find out that the electron's spin is anti-correlated with his positron's, but that's a totally legit, local quantum interaction. From my point of view, nothing happens faster than light. From my father's point of view, nothing happens faster than light. The only way to see anything happen faster than light is to be a third observer who can see what's happening in Boston and Philly simultaneously, which is impossible. As long as you restrict to what individual observers can see, no laws of physics are violated.

Of course, this sounded familiar. What was the solution to the black hole information-loss paradox? Realizing that no single observer can see both sides of a horizon. The solution to the apparent backward causation in top-down cosmology? Realizing that no single observer can be outside the universe to see causality violations. The solution to the firewall paradox? Realizing that no single observer can witness polygamous entanglement. Now the solution to the EPR paradox? Realizing that no single observer can see both measurements simultaneously. I couldn't be sure, but I was sensing a pattern here.

* * *

I called Rovelli at his home in Marseilles, in the south of France.

“Would it be fair to sum up the situation by saying that the interpretational difficulties in quantum mechanics come from trying to describe the world from an impossible view from nowhere?” I asked.

“Yes,” Rovelli agreed enthusiastically. “If we can renounce the view from nowhere, the view from the outside, and accept the idea of referring to observers, then the difficulties all go away. That's what I think.”

“Okay, so you always have to restrict to one frame of reference at a time, and the quantum weirdness arises when you compare your measurements to someone else's. But what about interference, like in the double-slit experiment? How is it that within a single reference frame you can see an interference pattern?”

“I like this question,” Rovelli said. “It forces thinking! When we say
interference
, what do we mean? In order to have interference, we need two of something to interfere. In the double-slit experiment, the interference is between the component of the electron passed through one slit and the component that passed through the other slit. But there is no electron passing through any slit in nature. This is just our language to mean ‘If I was at the slit and was measuring, I would see the electron here or there.'
Interference
is not a term denoting what is actually happening. It is a term referring to a comparison between what is observed by one observer and what would be observed by another observer. In your language, it is a term denoting the comparison between two different reference frames.”

That made sense. Interference was the interference of phases, and phases were just reference frames. Points of view. The strangeness of superposition wasn't the strangeness of many
worlds.
It was the strangeness of many
frames.

“As I've been looking at a number of different developments across physics, I'm coming to the realization that you always run into trouble when you try to describe physics from an impossible God's-eye view,” I said, “and that for anything in physics to make sense you have to define it in terms of a single observer's reference frame. It's starting to seem like there is one universe per observer. Like there's no way to talk about
the
universe.”

“I understand what you mean,” Rovelli said. “This is the challenging
aspect of modern physics. It forces us to renounce the previous image of a clear, objective, well-defined, perfectly describable state of affairs of the world. Quantum mechanics has asked us to give this up. This does come with a heavy metaphysical burden. Are we ready to rethink the universe in these terms? I think we have to take our physics seriously.”

“It's interesting that a lot of the progress physicists have made in understanding reality has come from accepting that more and more things are relative,” I mused, thinking of all the invariants that had slipped right off the IHOP napkin.

“Exactly!” Rovelli replied excitedly. “Exactly. When people heard that the Earth was round, it was very complicated conceptually to accept it. How could the people in Sydney be walking upside down? Then eventually people understood, there's no real up and down; they are relative. And they got used to it. Then it was hard to understand that motion was relative. Then it was hard to understand that simultaneity was relative. And I think quantum mechanics is a step in the same direction. It's telling us that the world is even more relative than expected. If one observer sees spin up, that doesn't necessarily mean that it's fixed forever for everyone else.”

Not fixed for everyone else. I couldn't help but think of Wheeler.
How preposterous to think that each has to invent the universe afresh.
“Did John Wheeler ever come across your work on relational quantum mechanics?” I asked. “I just spent time reading through his journals, and for years he seemed to be struggling with the question of multiple observers in quantum mechanics. He would ask things like, ‘Why does my measurement of the electron's spin fix it for every other observer?' ”

“Absolutely, yes,” Rovelli said. “I had a very warm relationship with John Wheeler. He got interested first in my work on quantum gravity. He sent me a very enthusiastic letter in his typical flowered style, which is today hanging on the wall of my office, inviting me to Princeton to deliver one of the first talks on loop quantum gravity—which, needless to say, raised unhappy criticisms from Ed Witten and David Gross, who were sitting uncomfortably in the audience.”

Criticism? From David Gross? I couldn't imagine it.

“Later, when I wrote the relational quantum mechanics paper, he reacted again enthusiastically and sent me a beautiful letter and a package of everything he had written on the argument, all tied into a handmade folder with an orange cover and a picture of his preferred ‘it from bit' image on the cover. My relational paper owes a lot to Wheeler's intuition, obviously. It was Wheeler who understood that information had to be the right concept. But he was always thinking about
the
observer, observing the world and making it happen and being part of it. He was focused on the circular path that this structure generated, and did not find a way to make sense of the relation between observers. I think he did appreciate my opening ‘the observer' into a multiplicity of observers, with any physical system being one, and especially the analysis of the coherence that quantum theory gives to the information of different observers. But he was aged when my paper came out, and I don't think he wrote anything on it.