Тед Чан - Exhalation - Stories

Set up a rack of billiard balls and execute a flawless break. Imagine the table has no pockets and is frictionless, so the balls just keep rebounding, never coming to a stop; how accurately can you predict the path of any given ball as it collides against the others? In 1978, the physicist Michael Berry calculated that you could predict only nine collisions before you would need to account for the gravitational effect of a person standing in the room. If your initial measurement of a ball’s position is off by even a nanometer, your prediction becomes useless within a matter of seconds.

The collisions between air molecules are similarly contingent and can be affected by the gravitational effect of a single atom a meter away. So even though the interior of a prism is shielded from the external environment, the result of the quantum measurement that takes place when the prism is activated can still exert an effect on the outside world, determining whether two oxygen molecules collide or whether they drift past each other. Without anyone intending it, the activation of the prism inevitably gives rise to a difference between the two branches generated. The difference is imperceptible at first, a discrepancy at the level of the thermal motion of molecules, but when air is turbulent, it takes roughly a minute for a perturbation at the microscopic level to become macroscopic, affecting eddies one centimeter in diameter.

For small-scale atmospheric phenomena, the effects of perturbations double in size every couple of hours. In terms of prediction, that means that an error one meter wide in your initial measurements of the atmosphere will lead to an error a kilometer wide in your prediction of the weather on the following day. At larger scales, the propagation of errors slows down due to factors like topography and the stratification of the atmosphere, but it doesn’t stop; eventually errors on the kilometer scale become errors hundreds or thousands of kilometers in size. Even if your initial measurements were so detailed that they included data about every cubic meter of the Earth’s atmosphere, your prediction of the future weather would cease to be useful within a month’s time. Increasing the resolution of the initial measurements has a limited benefit; because errors propagate so rapidly at the small scale, starting with data about every cubic centimeter of the atmosphere would prolong the accuracy of the prediction by only a matter of hours.

The growth of errors in weather prediction is identical to the divergence between the weather in the branches on opposite sides of a prism. The initial perturbation is the difference in the collision of oxygen molecules when the prism is activated, and within a month, the weather around the globe is different. Silitonga confirmed this when he and his parallel self exchanged weather reports one month after activating a prism. The weather reports were all seasonally appropriate—there was no location that experienced winter in one branch and summer in the other—but beyond that they were essentially uncorrelated. Without anyone making an effort, the two branches had diverged visibly on a worldwide scale.

After Silitonga published these results, in a paper titled “Studying Atmospheric Upscale Error Propagation with the Plaga Interworld-Signaling Mechanism,” historians engaged in heated debates over the extent to which weather could affect the course of history. Skeptics acknowledged that it could affect individuals’ daily lives in various ways, but how often were the outcomes of history-making events decided by the weather? Silitonga didn’t participate in the debates; he was waiting for his other, yearlong prism experiment to conclude.

There were times when the clients came in just the right order, and Wednesday afternoons were like that for Dana. The afternoon began with one of her most demanding clients, a man who asked her to make all his decisions for him, whined when she wouldn’t, and blamed her whenever he eventually did take an action. So it was a relief to see Jorge immediately afterward, a breath of fresh air to clear out her office. The issues he was dealing with weren’t the most interesting she’d ever seen, but she liked having him as a client. Jorge was funny and kind, and always well-intentioned; he was tentative about the therapeutic process, but they’d been making steady progress on his poor self-image and the negative attitudes that were holding him back.

Four weeks ago there had been an incident. Jorge’s manager at work was a mean-spirited tyrant who belittled everyone who worked for him; one of the ongoing themes of Dana’s sessions with Jorge was helping him to ignore his manager’s insults. One day, Jorge had lost his temper and punctured all four tires of his manager’s car when he was alone in the parking lot. Enough time had passed that it seemed like there was no risk of him getting caught, and while part of him wanted to pretend that it had never happened, part of him still felt terrible about what he’d done.

They began their session with some small talk; Dana got the sense that Jorge had something he wanted to say. She looked at him expectantly, and he said, “After our session last week, I went to one of those prism brokers, Lydoscope.”

Dana was surprised. “Really? What for?”

“I wanted to see how many versions of me acted the same way I did.”

“Tell me more.”

“I asked them to send questions to six versions of me. Since it’s such a recent departure point, it was cheap, so I asked for video. This morning they sent me a bunch of video files, recordings of what my paraselves said.”

“And what did you learn?”

“None of my paraselves have punctured their manager’s tires. All of them said they’ve fantasized about it. One came really close on the same day that I did it, but he stopped himself.”

“What do you think that means?”

“It means that my puncturing his tires was a freak accident. The fact that I did it doesn’t say anything important about me as a person.”

Dana knew of people using prisms in a similar way, but it was usually someone justifying their actions by pointing out they might have done something worse. She hadn’t encountered this particular version of it before, where the defense was based on their parallel selves behaving better. She certainly hadn’t expected it from Jorge. “So you think your paraselves’ behavior is a reflection on you?”

“The branches they checked, they were all ones where the departure point was just a month before the incident. That means that those paraselves were just the same as me; they hadn’t had time to become different people.”

She nodded; he was right about that. “Do you think the fact that you vandalized your manager’s car is canceled out by the fact that your paraselves didn’t?”

“Not canceled out, but it’s an indicator of the type of person I am. If all of my paraselves had punctured his tires, that would indicate something significant about my personality. That’s something Sharon would need to know about.” Jorge hadn’t told his wife about what he’d done; he’d been too ashamed. “But the fact that they didn’t means that I’m fundamentally not a violent person, so telling Sharon about what happened would give her the wrong idea.”

Getting him to tell his wife everything was something they’d have to build up to. “So how do you feel, now that you’ve gotten this information?”

“Relief, I suppose,” said Jorge. “I was worried about what it meant that I had done that. But now I’m not so worried.”

“Tell me more about that feeling of relief.”

“I feel like…” Jorge fidgeted in his chair as he searched for the words. Eventually he said, “I guess I feel like I got the results of a medical test back, and I’m in the clear.”