Will We Travel Back (Or Forward) In Time?

By J. Richard Gott Iii

We can travel through the three dimensions of space pretty much at will--moving forward or back, left or right, up or down--without even thinking about it. When it comes to the fourth dimension, though, we appear to be stuck. Time flows on in one direction only, and we flow with it like corks bobbing helplessly in a river. So the idea of traveling through time, as opposed to with time, is immensely seductive. Who wouldn't want to know what technology will look like in the year 3000, or witness the assassination of Julius Caesar?

Not only does such a thing seem extremely difficult, but it could also be a little risky. What if you prevented Caesar's assassination and changed history? What if you accidentally killed someone who happened to be your own ancestor? Then you wouldn't have been born, and couldn't have killed your ancestor, so you could be born after all to go back and...well, you get the idea.

We physicists are mindful of all these difficulties, of course. But we can't resist exploring the notion of time travel--not necessarily for practical reasons, but to understand the limits of our theories.

Do the laws of physics permit time travel, even in principle? They may in the subatomic world. A positron (the antiparticle associated with the electron) can be considered to be an electron going backward in time. Thus, if we create an electron-positron pair and the positron later annihilates in a collision with another, different electron, we could view this as a single electron executing a zigzag, N-shaped path through time: forward in time as an electron, then backward in time as a positron, then forward in time again as an electron.

The probability of a macroscopic object--like a human--doing this trick is infinitesimal. But thanks to Albert Einstein we know that time travel of a different sort does happen in the macroscopic world. As he showed back in 1905 with his special theory of relativity, time slows down for objects moving close to the speed of light, at least from the viewpoint of a stationary observer. You want to visit the earth 1,000 years from now? Just travel to a star 500 light-years away and return, going both ways at 99.995% the speed of light. When you return, the earth will be 1,000 years older, but you'll have aged only 10 years. I already know a time traveler. My friend, astronaut Story Musgrave, who helped repair the Hubble Space Telescope, spent 53.4 days in orbit. He is thus more than a millisecond younger than he would have been if he had stayed home. The effect is small, because he traveled very slowly relative to the speed of light, but it's real.

With more money, we could do better in the next century--but only a little. If we sent an astronaut to the planet Mercury and she lived there for 30 years before returning, she would be about 22 seconds younger than if she had stayed on Earth. Clocks on Mercury tick more slowly than those on Earth because Mercury circles the sun at a faster speed (and also because Mercury is deeper in the sun's gravitational field; gravity affects clocks much as velocity does). Astronauts traveling away from Earth to a distance of 0.1 light years and returning at 1% the speed of light would arrive back 8.8 hours younger than if they hadn't gone.

The downside of traveling into the future this way is that you might be stuck there. Is there any way of going backward in time? Once again, Einstein may have provided the answer. His 1915 theory of general relativity showed that space and time are curved, and that the curvature can be large in the neighborhood of very massive objects. If an object is dense enough, the curvature can become nearly infinite, perhaps opening a tunnel that connects distant regions of space-time as though they were next door. Physicists call this tunnel a wormhole, in an analogy to the shortcut a worm eats from one side of a curved apple to the other.

In 1988, Kip Thorne, a physicist at Caltech, and several colleagues suggested that you could use such a wormhole to travel into the past. Here's how you do it: move one mouth of the wormhole through space at nearly the speed of light while leaving the other one fixed. Then jump in through the moving end. Like a moving astronaut, this end ages less, so it connects back to an earlier time on the fixed end. When you pop out through the fixed end an instant later, you'll find that you've emerged in your own past.

The problem with wormholes is that the openings are microscopic and tend to snap shut a fraction of a second after they're created. The only way to keep them open, as far as we know, is with matter that has negative density. In layman's terms, that's stuff that weighs less than nothing. This may sound impossible, but the Dutch physicist Hendrik Casimir theorized in 1948 that holding two plates of electrically conducting material very close together in a vacuum actually does create a region of negative density that exerts an inward pressure on the plates. The force predicted by Casimir has been verified in the laboratory.

Using this idea, Thorne and his colleagues proposed constructing a wormhole tunnel 600 million miles in circumference, with Casimir plates separated by only 400 proton diameters at the midpoint. Time travelers would have to somehow open doors in these plates to pass through the wormhole. The mass required for construction? Two hundred million times the mass of the sun. These are projects only a supercivilization could attempt--not something for 21st century engineers.

In 1991 I found another possible mechanism for time travel using cosmic strings, thin strands of energy millions of light-years long, predicted by some theories of particle physics (but not yet observed in the universe). You could try to construct a cosmic-string time machine by finding a large loop of cosmic string and somehow manipulating it so it would contract rapidly under its own tension, like a rubber band. The extraordinary energy density of the string curves space-time sharply, and by flying a spaceship around the two sides of the loop as they pass each other at nearly the speed of light, you'd travel into the past.

To go back in time by one year, unfortunately, you'd need a loop containing about half the mass-energy of an entire galaxy. Worse yet, the contracting cosmic-string loop would probably trigger the formation of a rotating black hole, trapping any time-travel regions inside. You would almost certainly be torn apart by near infinite space curvature before you could travel anywhere.

Even if you could get past these difficulties, the physics of both types of time machine dictate that you can't go back in time to an epoch before the time machine was created. So you couldn't meet and perhaps kill your own ancestor. But if such a machine were built today, your descendants might come and kill you, changing their own past.

Some argue conservatively that time travelers don't change the past; they were always part of it. On the other hand, paradoxical though this sounds, a version of the many-worlds theory of quantum mechanics (see "Will We Discover Another Universe?" in this issue) devised by Oxford physicist David Deutsch might allow such history-changing visits. In this picture, there are many interlacing world histories, so that if you went back in time and killed your grandmother when she was a young girl, this would simply cause space-time to branch off into a new parallel universe that doesn't interfere with the familiar one.

Stephen Hawking has addressed the problem in a different way, proposing what he calls a chronology-protection conjecture. Somehow, he argues, the laws of physics must always conspire to prevent travel into the past. He believes that quantum effects, coupled with other constraints, will always step in to prevent time machines. The jury is still out on this question. We may need to develop a theory of quantum gravity to learn whether Hawking is right.

So, will we time-travel in the next century? Travel to the future--yes, but only in short hops, I suspect. To the past--very likely not. Such travel is expensive, dangerous and subject to quantum effects that may or may not spoil your chances of coming back alive. Those of us working in this field aren't rushing to the patent office with time-machine blueprints. But we are interested in knowing whether time machines are possible, even in principle, because answering that question will tell us where the boundaries of physics lie and provide clues to how the universe works.

J. Richard Gott III is a professor of astrophysics at Princeton, where he does research on general relativity and cosmology