That belief crumbled when experiments began to reveal the indeterministic effects of quantum mechanics—for instance, in the radioactive decay of atoms. The problem goes like this, Tollaksen says: Take two radioactive atoms, so identical that “even God couldn’t see the difference between them.” Then wait. The first atom might decay a minute later, but the second might go another hour before decaying. This is not just a thought experiment; it can really be seen in the laboratory. There is nothing to explain the different behaviors of the two atoms, no way to predict when they will decay by looking at their history, and—seemingly—no definitive cause that produces these effects. This indeterminism, along with the ambiguity inherent in the uncertainty principle, famously rankled Einstein, who fumed that God doesn’t play dice with the universe.

It bothered Aharonov as well. “I asked, what does God gain by playing dice?” he says. Aharonov accepted that a particle’s past does not contain enough information to fully predict its fate, but he wondered, if the information is not in its past, where could it be? After all, something must regulate the particle’s behavior. His answer—which seems inspired and insane in equal measure—was that we cannot perceive the information that controls the particle’s present behavior because it does not yet exist.

“Nature is trying to tell us that there is a difference between two seemingly identical particles with different fates, but that difference can only be found in the future,” he says. If we’re willing to unshackle our minds from our preconceived view that time moves in only one direction, he argues, then it is entirely possible to set up a deterministic theory of quantum mechanics.

In 1964 Aharonov and his colleagues Peter Bergmann and Joel Lebowitz, all then at Yeshiva University in New York, proposed a new framework called time-symmetric quantum mechanics. It could produce all the same treats as the standard form of quantum mechanics that everyone knew and loved, with the added benefit of explaining how information from the future could fill in the indeterministic gaps in the present. But while many of Aharonov’s colleagues conceded that the idea was built on elegant mathematics, its philosophical implications were hard to swallow. “Each time I came up with a new idea about time, people thought that something must be wrong,” he says.

Perhaps because of the cognitive dissonance the idea engendered, time-symmetric quantum mechanics did not catch on. “For a long time, it was nothing more than a curiosity for a few philosophers to discuss,” says Sandu Popescu at the University of Bristol, in England, who works on the time-symmetric approach with Aharonov. Clearly Aharonov needed concrete experiments to demonstrate that actions carried out in the future could have repercussions in the here and now.

Through the 1980s and 1990s, Tollaksen teamed up with Aharonov to design such upside-down experiments, in which outcome was determined by events occurring after the experiment was done. Generally the protocol included three steps: a “preselection” measurement carried out on a group of particles; an intermediate measurement; and a final, “postselection” step in which researchers picked out a subset of those particles on which to perform a third, related measurement. To find evidence of backward causality—information flowing from the future to the past—the experiment would have to demonstrate that the effects measured at the intermediate step were linked to actions carried out on the subset of particles at a later time.

Tollaksen and Aharonov proposed analyzing changes in a quantum property called spin, roughly analogous to the spin of a ball but with some important differences. In the quantum world, a particle can spin only two ways, up or down, with each direction assigned a fixed value (for instance, 1 or –1). First the physicists would measure spin in a set of particles at 2 p.m. and again at 2:30 p.m. Then on another day they would repeat the two tests, but also measure a subset of the particles a third time, at 3 p.m. If the predictions of backward causality were correct, then for this last subset, the spin measurement conducted at 2:30 p.m. (the intermediate time) would be dramatically amplified. In other words, the spin measurements carried out at 2 p.m. and those carried out at 3 p.m. together would appear to cause an unexpected increase in the intensity of spins measured in between, at 2:30 p.m. The predictions seemed absurd, as ridiculous as claiming that you could measure the position of a dolphin off the Atlantic coast at 2 p.m. and again at 3 p.m., but that if you checked on its position at 2:30 p.m., you would find it in the middle of the Mediterranean.

And the amplification would not be restricted to spin; other quantum properties would be dramatically increased to bizarrely high levels too. The idea was that ripples of the measurements carried out in the future could beat back to the present and combine with effects from the past, like waves combining and peaking below a boat, setting it rocking on the rough sea. The smaller the subsample chosen for the last measurement, the more dramatic the effects at intermediate times should be, according to Aharonov’s math. It would be hard to account for such huge amplifications in conventional physics.

For years this prediction was more philosophical than physical because it did not seem possible to perform the suggested experiments. All the team’s proposed tests hinged on being able to make measurements of the quantum system at some intermediate time; but the physics books said that doing so would destroy the quantum properties of the system before the final, postselection step could be carried out. Any attempt to measure the system would collapse its delicate quantum state, just as chasing dolphins in a boat would affect their behavior. Use this kind of invasive, or strong, measurement to check on your system at an intermediate time, and you might as well take a hammer to your apparatus.

By the late 1980s, Aharonov had seen a way out: He could study the system using so-called weak measurements. (Weak measurements involve the same equipment and techniques as traditional ones, but the “knob” controlling the power of the observer’s apparatus is turned way down so as not to disturb the quantum properties in play.) In quantum physics, the weaker the measurement, the less precise it can be. Perform just one weak measurement on one particle and your results are next to useless. You may think that you have seen the required amplification, but you could just as easily dismiss it as noise or an error in your apparatus.

The way to get credible results, Tollaksen realized, was with persistence, not intensity. By 2002 physicists attuned to the potential of weak measurements were repeating their experiments thousands of times, hoping to build up a bank of data persuasively showing evidence of backward causality through the amplification effect.

Just last year, physicist John Howell and his team from the University of Rochester reported success. In the Rochester setup, laser light was measured and then shunted through a beam splitter. Part of the beam passed right through the mechanism, and part bounced off a mirror that moved ever so slightly, due to a motor to which it was attached. The team used weak measurements to detect the deflection of the reflected laser light and thus to determine how much the motorized mirror had moved.

That is the straightforward part. Searching for backward causality required looking at the impact of the final measurement and adding the time twist. In the Rochester experiment, after the laser beams left the mirrors, they passed through one of two gates, where they could be measured again—or not. If the experimenters chose not to carry out that final measurement, then the deflected angles measured in the intermediate phase were boringly tiny. But if they performed the final, postselection step, the results were dramatically different. When the physicists chose to record the laser light emerging from one of the gates, then the light traversing that route, alone, ended up with deflection angles amplified by a factor of more than 100 in the intermediate measurement step. Somehow the later decision appeared to affect the outcome of the weak, intermediate measurements, even though they were made at an earlier time.

This amazing result confirmed a similar finding reported a year earlier by physicists Onur Hosten and Paul Kwiat at the University of Illinois at Urbana-Champaign. They had achieved an even larger laser amplification, by a factor of 10,000, when using weak measurements to detect a shift in a beam of polarized light moving between air and glass.

For Aharonov, who has been pushing the idea of backward causality for four decades, the experimental vindication might seem like a time to pop champagne corks, but that is not his style. “I wasn’t surprised; it was what I expected,” he says.

Paul Davies, a cosmologist at Arizona State University in Tempe, admires the fact that Aharonov’s team has always striven to verify its claims experimentally. “This isn’t airy-fairy philosophy—these are real experiments,” he says. Davies has now joined forces with the group to investigate the framework’s implications for the origin of the cosmos (See “Does the Universe Have a Destiny?” below).

Vlatko Vedral, a quantum physicist at the University of Oxford, agrees that the experiments confirm the existence and power of weak measurements. But while the mathematics of the team’s framework offers a valid explanation for the experimental results, Vedral believes these results alone will not be enough to persuade most physicists to buy into the full time-twisting logic behind it.

For Tollaksen, though, the results are awe-inspiring and a bit scary. “It is upsetting philosophically,” he concedes. “All these experiments change the way that I relate to time, the way I experience myself.” The results have led him to wrestle with the idea that the future is set. If the universe has a destiny that is already written, do we really have a free choice in our actions? Or are all our choices predetermined to fit the universe’s script, giving us only the illusion of free will?

Tollaksen ponders the philosophical dilemma. Was he always destined to become a physicist? If so, are his scientific achievements less impressive because he never had any choice other than to succeed in this career? If I time-traveled back from the 21st century to the shores of Lake Michigan where Tollaksen’s 13-year-old self was reading the works of Feynman and told him that in the future I met him in the Azores and his fate was set, could his teenage self—just to spite me—choose to run off and join the circus or become a sailor instead?

The free will issue is something that Tollaksen has been tackling mathematically with Popescu. The framework does not actually suggest that people could time-travel to the past, but it does allow a concrete test of whether it is possible to rewrite history. The Rochester experiments seem to demonstrate that actions carried out in the future—in the final, postselection step—ripple back in time to influence and amplify the results measured in the earlier, intermediate step. Does this mean that when the intermediate step is carried out, the future is set and the experimenter has no choice but to perform the later, postselection measurement? It seems not. Even in instances where the final step is abandoned, Tollaksen has found, the intermediate weak measurement remains amplified, though now with no future cause to explain its magnitude at all.

I put it to Tollaksen straight: This finding seems to make a mockery of everything we have discussed so far.

Tollaksen is smiling; this is clearly an argument he has been through many times. The result of that single experiment may be the same, he explains, but remember, the power of weak measurements lies in their repetition. No single measurement can ever be taken alone to convey any meaning about the state of reality. Their inherent error is too large. “Your pointer will still read an amplified result, but now you cannot interpret it as having been caused by anything other than noise or a blip in the apparatus,” he says.

In other words, you can see the effects of the future on the past only after carrying out millions of repeat experiments and tallying up the results to produce a meaningful pattern. Focus on any single one of them and try to cheat it, and you are left with a very strange-looking result—an amplification with no cause—but its meaning vanishes. You simply have to put it down to a random error in your apparatus. You win back your free will in the sense that if you actually attempt to defy the future, you will find that it can never force you to carry out postselection experiments against your wishes. The math, Tollaksen says, backs him on this interpretation: The error range in single intermediate weak measurements that are not followed up by the required post­selection will always be just enough to dismiss the bizarre result as a mistake.

Tollaksen sums up this confounding argument with one of his favorite quotes, from the ancient Jewish sage Rabbi Akiva: “All is foreseen; but freedom of choice is given.” Or as Tollaksen puts it, “I can have my cake and eat it too.” He laughs.

Here, finally, is the answer to Aharonov’s opening question: What does God gain by playing dice with the universe? Why must the quantum world always retain a degree of fuzziness when we try to look at it through the time slice of the present? That loophole is needed so that the future can exert an overall pull on the present, without ever being caught in the act of doing it in any particular instance.

“The future can only affect the present if there is room to write its influence off as a mistake,” Aharonov says.

Whether this realization is a masterstroke of genius that explains the mechanism for backward causality or an admission that the future’s influence on the past can never fully be proven is open to debate. Andrew Jordan, who designed the Rochester laser amplification experiment with Howell, notes that there is even fundamental controversy over whether his results support Aharonov’s version of backward causality. No one disputes his team’s straightforward experimental results, but “there is much philosophical thought about what weak values really mean, what they physically correspond to—if they even really physically correspond to anything at all,” Jordan says. “My view is that we don’t have to interpret them as a consequence of the future’s influencing the present, but rather they show us that there is a lot about quantum mechanics that we still have to understand.” Nonetheless, he is open to being convinced otherwise: “A year from now, I may well change my mind.”

Popescu argues that the Rochester findings are hugely important because they open the door to a completely new range of laboratory explorations based on weak measurements. In starting from the conventional interpretation of quantum mechanics, physicists had not realized such measurements were possible. “With his work on weak measurements, Aharonov began to pose questions about what is possible in quantum mechanics that nobody had ever even thought could be articulated,” Popescu says.

Aharonov remains circumspect. He has spent most of his adult life waiting for recognition of the merit of his theory. If it is destined that mainstream physics should finally take serious notice of his time-twisting ideas, then so it will be.

And Tollaksen? He too is at one with his destiny. A few months ago he moved to Laguna Beach, California. “I’m in a house where I can hear the surf again—what a relief,” he says. He feels that he is finally back to where he was always meant to be.

DOES THE UNIVERSE HAVE A DESTINY?

Is feedback from the future guiding the development of life, the universe, and, well, everything? Paul Davies at Arizona State University in Tempe and his colleagues are investigating whether the universe has a destiny—and if so, whether there is a way to detect its eerie influence.

Cosmologists have long been puzzled about why the conditions of our universe—for example, its rate of expansion—provide the ideal breeding ground for galaxies, stars, and planets. If you rolled the dice to create a universe, odds are that you would not get one as handily conducive to life as ours is. Even if you could take life for granted, it’s not clear that 14 billion years is enough time for it to evolve by chance. But if the final state of the universe is set and is reaching back in time to influence the early universe, it could amplify the chances of life’s emergence.

With Alonso Botero at the University of the Andes in Colombia, Davies has used mathematical modeling to show that bookending the universe with particular initial and final states affects the types of particles created in between. “We’ve done this for a simplified, one-dimensional universe, and now we plan to move up to three dimensions,” Davies says. He and Botero are also searching for signatures that the final state of the universe could retroactively leave on the relic radiation of the Big Bang, which could be picked up by the Planck satellite launched last year.

Ideally, Davies and Botero hope to find a single cosmic destiny that can explain three major cosmological enigmas. The first mystery is why the expansion of the universe is currently speeding up; the second is why some cosmic rays appear to have energies higher than the bounds of normal physics allow; and the third is how galaxies acquired their magnetic fields. “The goal is to find out whether Mother Nature has been doing her own postselections, causing these unexpected effects to appear,” Davies says.

Bill Unruh of the University of British Columbia in Vancouver, a leading physicist, is intrigued by Davies’s idea. “This could have real implications for whatever the universe was like in its early history,” he says.