Even though the laws of physics seem to permit time travel, many physicists and non-physicists still worry about the paradoxes that arise when you try to change the past (for an example see… well, pretty much any movie involving time travel). But quantum mechanics can give a way out of the mess, and an experiment performed at the University of Queensland has tested what happens when you try this.

Although not a dinkum time machine, the experiment simulated the effect of sending a photon (a particle of light) back in time by using two photons, one acting as the past version and one as the future version, and having them interact (Martin Ringbauer, Matthew A. Broome, Casey R. Myers, Andrew G. White & Timothy C. Ralph 2014, “Experimental simulation of closed timelike curves”, *Nature Communications* **5**, Article number: 4145, doi:10.1038/ncomms5145).

To make it act like a time machine, they imposed theoretical conditions established in 1991 by David Deutsch (Deutsch 1991, “Quantum mechanics near closed timelike lines”, *Physical Review D* **44**, 3197, DOI:10.1103/PhysRevD.44.3197 [PDF 4.7 MB]).

In Deutsch’s theory, any attempt to meddle with the past gives mixed results, i.e. a quantum mixture of meddled and non-meddled. These separate possibilities exist simultaneously, and are often interpreted as alternative timelines created by the act of travelling back in time.

There are other theories of time travel, such as those that forbid any changes to the past that haven’t already happened, but Deutsch’s model is a particularly popular one.

The study of even a simulated time machine gives clues to how our understanding of physics may have to change to accommodate such bizarre circumstances.

To find out more about it, we spoke to one of the experimenters, PhD student Martin Ringbauer.

**Q: Now, I guess the big question is, have you actually achieved time travel? Are these photons really going back in time.**

Well, the photons are definitely not going back in time actually. As you can imagine, that’s not something that can happen in our everyday life here.

So, to give a bit of background, it turns out that time travel seems to be possible in general relativity. The mathematical formalism tells us that there can be something like a closed timelike curve, which is a path in spacetime that goes back to the same point in space and time. So this would lead to time travel.

Now, there has been a couple of solutions found—solutions to Einstein’s field equations—that involve such closed timelike curves. There is for example, big rotating dust distributions, or black holes, or traversable wormholes. But all of them are extreme effects, so there is some really big gravitational fields, so this is nothing that we can have in the lab. But at least it’s unlikely that we see something like this on Earth.

So what we did, is we simulated that. What we wanted to look at is what happens if a photon—so it’s a particle of light—goes back in time, interacts with itself, and how does such a system behave?

**Q: I imagine there’s some sort of magical crystals involved in doing all of this?**

Yes, so the photons we create are created by a spontaneous parametric down-conversion, which basically means you shine a strong laser pulse onto a non-linear crystal, and the crystal then creates pairs of photons. That’s single photons, so we can then use them in our experiment.

The important point is that we get them in pairs, so we can use one of them, for example, to trigger the presence of the other one. So we always know when there is a photon.

And we use those photons in our experiment then, so we wanted to simulate the evolution of one photon, but to do this we use two photons. Yeah, we don’t have a photon that actually travels back in time, so we use the second photon as the past incarnation of this first photon.

**Q: OK, so you basically have two identical photons, one is the future photon and one is the past photon, and you make them interact with each other. **

Yes, they are not quite identical, because our whole simulation is built on Deutsch’s theory, that was developed in 1991, which is a self-consistent formulation of time travel, of closed timelike curves, for quantum systems. It doesn’t work for classical systems, but we can talk about that a bit later.

But for quantum systems, it gives us a self-consistent formalism, and the central element is the consistency relation, which basically is the assumption that whatever goes into the time machine comes out again. And this is what we impose in our simulation: this is how we simulate this evolution.

We impose the consistency relation on this second photon, which is the past version, and that way we get consistent overall evolution of this one photon that we’re studying.

**Q: So there are many different ideas of what happens if you were to try and go back and change the past, and I suppose people either think that you can’t change what’s already happened, or that if you do try you create an alternative timeline. What sort of model is your experiment in line with?**

An important point is first of all that in any classical case, time travel immediately leads to paradoxes, like the famous Grandfather Paradox. You go back in time, you kill your grandfather, that means you were never born so you could not have gone back in time, so you should be alive, and it goes on like this.

In the quantum case, that is what I meant by self-consistent formalism, these paradoxes can be resolved and you can formulate time travel in a consistent way. Which is basically a consequence of the fact that the quantum systems can be in mixtures of states.

So if you map the Grandfather Paradox to a photon, then you could say whenever the photon is in state 1, that corresponds to the existence of the time traveller, and when it’s in state 0, it corresponds to non-existence.

So whenever a photon in state 1 meets another photon in state 1, it would be flipped to 0. That is the Grandfather Paradox: when you meet your grandfather you kill him.

Now in the quantum case, you can not only exist in state 0 or 1, but also in a mixture of those states. And this is what resolves the paradox. So if you design a situation where the Grandfather Paradox would arise in a quantum case, then you just end up in a mixed state of existing and non-existing.

Which is fine for quantum systems, but it doesn’t work for classical systems. That’s why we cannot resolve the paradox there in the same way.

**Q: OK, but again this isn’t real time travel we’re talking about here, so why is it important to test this sort of thing? Does this tell you something about how quantum physics works?**

That tells us something about where our problems are at least. So, quantum physics works very well, same as general relativity, they’re both very well-tested theories that describe our world very well in their regimes. So general relativity works well for very large systems, like stars, galaxies; quantum mechanics for very small, like light particles.

But they don’t really fit together, and in particular we don’t really know what happens when we look at quantum mechanics in the presence of strong gravitational fields. And this is where our simulation comes in, because you would need very strong gravitational fields to have closed timelike curves.

And then it seems like they are consistent with both quantum mechanics and general relativity, but what happens is that the laws of quantum mechanics seem to change quite a bit. So there are things happening, like perfect cloning of states, perfect discrimination of quantum states, which is not possible normally in quantum mechanics.

These are the effects that our study pointed out and that we want to study and understand. We hope that that helps to understand where the problems are in reconciling those two big theories.

**Q: Well, it’s quite a good step along that path to answer some fundamental questions. It must be great for you to see your work have such an impact, and I think congratulations are in order for getting published in Nature Communications. How’s the rest of the media attention been with this though?**

As it is with the topic of time travel it’s easily over-interpreted.

In particular, we cannot make any statement about what happens in the classical case, because there are still those paradoxes and our study does not address that. And we also don’t build a time machine or anything.

But what we do is we study what happens, what are the consequences of this consistent model of time travel that is there. And we don’t have any good physical reason to believe that closed timelike curves don’t exist. The only reason that we have—the only argument against it that we have—is those paradoxes in the classical case, which disappear in the quantum case and can be resolved.

**Q: This sounds like the beginning of a long and distinguished career for you. I hope that soon we will hear about you actually going back in time for real. Good luck with the PhD though, and thank you very much for talking to us.**

It was a pleasure.

(This story went to air on 10 July 2014. You can listen to the podcast.)