The physics of time travel

The majority of physicists say time travel probably won't work (at least in the Hollywood-go-anytime-you-wanna sense). Several time traveler parties have gone famously unattended by time travelers (at least, any willing to fess up about it). In general, science is kind of a buzz-kill on this one.

But if you want some justification for your daydreams, the person to talk to is Ronald Mallett, a theoretical physicists at the University of Connecticut who is most well-known for being the guy who thinks time travel is totally possible. (Mark wrote about him here back in 2007.) In fact, in 2006, Mallett predicted that time travel would be figured out within a decade.

I honestly have not researched this enough to give you my opinion on Mallett's ideas. His fellow physicists have addressed it, though. You can read one response to Mallett at arXiv. All of that is a long, context-relevant introduction to the video above, where Mallett explains his theories. I wanted to post the video because it's interesting and I thought you all would dig it. I'm also interested in the new video series this comes from—EPIPHANY, a daily video about big ideas taken from interviews with journalists, tech thinkers, scientists, and more. Mallett makes an interesting kick off for a series like this.

My hope, though, is that EPIPHANY doesn't only focus on scientific ideas that are kind of on the fringe. There's so many amazing discoveries that have the bulk of evidence behind them, it seems like a waste of a good platform to not cover the stuff that's more likely to be true.

Check out the rest of the EPIPHANY videos. So far, the series includes clips from interviews with Ronald Mallett, Reddit founder Alexis Ohanian, and filmmaker Jason Silva.


  1. “the person to talk to is Ronald Mallett, a theoretical physicists”
    Well sure, I’d assume he knows a bit about time travel if he’s already inhabiting multiple bodies.

    Man, when’s the Primer guy going to make another film?

    1. “Man, when’s the Primer guy going to make another film?”

      Apparently he’s got a new project in the works:

  2. Time travel is completely possible!

    It only works one way though, and you can only travel at the speed of time.

          1. Phase sensitivity: I believe the calculation can only be done using the log out of the south end of a northbound horse.

  3. Shucks, the Pictures for Sad Children archives are all mixed up.  Fortunately I could still find this somewhere:

  4. Re traveling forward: He says “when we have rockets that can go near the speed of light.” But: as velocity increases, so does mass; and as it turns out, to accelerate conventionally (ie, a rocket), you’d need near-infinite power. So much for traveling into the future with rocket power. Re traveling backward: He says that the universe can create “closed time-curves.” But the distances we’re talking about are huge, and to transit them in a reasonable fraction of a human lifetime, one would require near-infinite power, and a speed so near light (a low tau) that again, we’re talking about unreachable extremes.

    So, no. No significant time time travel. Not using those mechanisms, anyway.

    1. Mass does not increase in the rocket’s own frame of reference, so this doesn’t have an effect on the amount of fuel needed. However, constant acceleration for long periods of time requires huge amounts of fuel in both Newtonian physics and relativity, because the fuel used up in earlier stages of the journey has to accelerate not just the payload (crew and habitation) but also all the remainder of the fuel for the rest of the journey. It works out that fuel mass increases exponentially with the length of time you want to accelerate, see the bottom part of this page (the equation for the ratio of fuel mass to payload mass is e^(aT/c) – 1, where a is the magnitude of the constant acceleration, T is the time of the acceleration as measured by clocks onboard the ship, and c is the speed of light).

      What do you mean when you say “the distances we’re talking about are huge”? Are you talking about traveling on a closed timelike curve by traveling on an extended journey in a rotating Godel universe? (more info on rotating universes here) It’s true in this case the distances would be huge, but that’s not the only way to get a closed timelike curve in Einstein’s theory of general relativity, there is also the theoretical possibility of traversable wormholes. But although they could theoretically be used for time travel in pure general relativity, there are some hints that when quantum effects are considered, they might destroy the wormhole before it the two ends could be moved into the right positions to allow backwards time travel. Ultimately physicists can’t be sure if backwards time travel is allowed by the laws of physics until they have a full theory of “quantum gravity” like string theorists are hoping for (and I’ve read that there are also some hints from string theory that backwards time travel would be forbidden, see here).

      For anyone interested in a mostly equation-free introduction to the physics of time travel, from the perspective of a physicist who’s not as controversial as Ronald Mallett, I highly recommend Kip Thorne’s book Black Holes and Time Warps–Thorne is actually the one who discovered the theoretical possibility of traversable wormholes in Einstein’s theory (he did so when Carl Sagan asked him about a semi-plausible way for aliens to get around the light speed barrier for Sagan’s book Contact)

      1. I’ll second that recommendation of Black Holes & Time Warps – I used that book quite extensively for research. He talks about the possibility of developing time travel into the future and into the past, using wormholes, if a way to both open them and keep them open could ever be found, so long as the point you’re travelling to is within the lifetime of the wormhole, and all within the tenets of Einsteinian relativity.

      2. Wormholes: there is no way to practically make a stable wormhole. Not even a theory of how. So, again, no. Yes, we’re talking about transiting a Godel universe, because that doesn’t require us to create  constructs impossible to our technology. Unfortunately, as I said, it *does* create a “you can’t get there from here because you’ll be dead” problem.

        Mass: achieving a low tau is impossible. Not enough energy to apply.

        1. I agree there’s currently no known way to create wormholes, but the speculation is that when we understand quantum gravity better it might show a way, since quantum gravity probably will allow for spontaneous changes in the topology of spacetime (which would be required for the simplest type of wormhole). Obviously it’s speculative, but if we ignore Mallett’s far-fetched claims about time travel in the near future, I think most physicists agree even if it’s allowed in principle by the laws of physics, it would require a civilization with technology advanced vastly beyond our own to actually achieve. In any case I think this is more of a realistic possibility than the possibility we will be able to use Godel’s method in the future, since the observational evidence seems pretty strong that our universe isn’t rotating at the large rate required.

          Achieving a low tau or equivalently a large relativistic gamma factor (note that this use of “tau” isn’t quite standard terminology though it was used in the classic science fiction novel Tau Zero, so I prefer to use gamma) is not “impossible”, just very difficult and again requiring a pretty advanced civilization (though not quite as advanced as building wormholes, I think). With an engine of ideal efficiency in converting fuel mass to kinetic energy in the form of photons (like an ideal matter/antimatter drive), an equation at the bottom of the relativistic rocket page shows that if you accelerate continuously to your destination, the ratio of fuel mass to payload mass will be given by M/m = gamma*(1 + v/c) – 1. Since gamma = 1/sqrt[1 – v^2/c^2], v/c = sqrt[1 – (1/gamma^2)]. So for to achieve gamma=100 (at which point you would be aging 100 times slower than observers on Earth, as measured in Earth’s frame), you would need a ratio of fuel mass to payload mass of 100*(1 + sqrt[1 – (1/100^2)]) – 1, plugging that into the calculator here gives a final result of about 199 times more fuel mass than payload mass (if you want to calculate the fuel:payload needed for different values of gamma, just replace “100” in the above expression with some other number and copy and past the equation into the online calculator). Of course if you wanted to turn around and return to Earth more fuel would be needed, but our hypothetical advanced civilization might have a refueling outpost at the turnaround location, and it’s also been suggested that a bussard ramjet could be used for braking with respect to the interstellar medium (though the original idea of using a ramjet to continuously accelerate relative to the interstellar medium, as in Tau Zero, seems to be unworkable), so you wouldn’t necessarily need to carry fuel for slowing down. There are also some suggestions for how you could accelerate to relativistic speeds in the Earth’s frame without actually needing to carry fuel on board the craft, like the idea of a ship with a giant sail or pusher plate that is pushed along by the beam from a powerful laser in the solar system, or a beam of microscopic pellets accelerated to relativistic speed by something like a giant mass driver. Another idea is that the bussard ramjet could actually be used for increasing speed if it didn’t have to react with the natural interstellar medium, but instead could travel along a trajectory that had been pre-seeded with fuel pellets. See this post and this one at the “Centauri Dreams” blog for some details on these proposed methods, or pages 136-148 of the book Centauri Dreams (available on google books).

          Though these sorts of technologies are not as advanced as wormholes, still I would imagine that if we ever manage to achieve them, by that time “we” will no longer be biological humans but rather mind uploads or some other form of AI. And in that case, the whole idea of using relativistic velocities for “forward time travel” will be kind of a moot point, since a computer-based intelligence can achieve the same subjective effect just by slowing the speed of the simulation relative to realtime. Obviously relativistic travel would still be useful for exploring the universe in reasonable amounts of time, though.

  5. I’m sorry, but I have difficulty correlating the title ‘theoretical physicist’ with someone who thinks that GPS satellites are in geosynchronous orbit.  (Ah, got it, he’s theoretically a physicist, just isn’t one in practice)

  6. “Mallett predicted that time travel would be figured out within a decade.”

    Even if you reject the theory that all time (past, present and future) all exist at once.
    If time travel ever is figured out, it has already been figured out.

    1. If you think of time as just another dimension, then it makes sense to say all different times are equally “real”, but still distinguish between events that happen at one time and events that happen at another. Of course if time travel is invented at one point on the timeline then time travelers can travel to other points too, but there’d still be a single date when it was figured out, and the theories about time travel in modern physics generally say any kind of time machine (a wormhole, say) couldn’t be used to travel back to before the moment the time machine was first created.

  7. Either it’s not possible or it’s possible and we destroy ourselves from the mess that it would create. 

  8. The real problem is that time does not really exist. It is an arbitrary social construct, not a quantifiable force like gravity or magnetism.

    1. Do you think space exists? Even if the answer is no, you presumably know what people are talking about when they talk of “traveling through space”, even if you would say this is just a confused way of stating something else like “traveling from contact with one physical object to another”. If the answer is yes, I’d be curious why you think space is more real than time–they are both treated as dimensions in 4-dimensional spacetime in relativity, and in relativity there is no unique way to define the “present moment” (because of the relativity of simultaneity, two events which happen at the “same moment” in one observer’s frame of reference can happen at “different moments” in another observer’s frame, and physically both frames are considered equally valid, there is no reason to prefer any particular frame’s way of dividing up 4D spacetime into snapshots of 3D space at different times…see also the Rietdijk–Putnam argument)

    2. Yes, or to put it another way, the biggest mistake made was to call time a dimension.
      In a black hole or in between the Galaxies there is only “now”. Time dilation, yes. We can only measure ‘time’ by mechanical means. The swing of a pendulum, oscillation of crystal, atomic clocks. All ONLY illustrate the rate at which physical processes ‘occur’ or run, if you must, in a particular inertial frame.
      No evidence of wormholes, only speculation derived from mathematics.
      They will be saying neutrinos travel faster than the speed of light next!

      1. Like I said above, if you want to say “there is only now” (the presentist philosophical view of time, as opposed to the alternative of eternalism) you have to contend with the relativity of simultaneity–are you familiar with this idea? If you understand that different reference frames define simultaneity (and thus “now”) differently, do you also understand that the laws of physics work exactly the same when measured in these different frames, so there is no physical basis for deciding which frame’s definition is “correct”? And are you assuming future discoveries in physics will break this symmetry and discover a “preferred frame” for the fundamental laws of physics (something I think the majority of physicists would see as very unlikely), or are you OK with the idea that the truth about simultaneity might be completely impossible to determine empirically, even in principle? So that like the existence of heaven or other postulated nonphysical realities, the existence of “true simultaneity” would be something that believers just have to take on faith?

        1. Thank you, but the problem with simultaneity is that the speed of light does not allow events to be compared.  Since events in one frame of reference cannot be compared with events in others does not rule out a common ‘Now’.
          A mistake, which has a just little reference to our discussion, is the assumption of a ‘time line’  shortly after the ‘Big bang’. Fractions of seconds are quoted in the instants when ‘time’ would have been impossible to assess, both practically and theoretically,  since the measurement of time is relative to the adjacent mass. (Refraction of a photon passing the Sun). The Michelson Morley experimental results have components of this, in as much as the speed of light is only relative to adjacent Mass, as near a Black Hole. The experiment was originally conceived to investigate Aether drag, but it has the greater value when looked at from the point of view that a component of Matter must carry the Electromagnetic wave. We must remember that ‘particles’  do not exist at a point. They may carry electrostatic and gravitational ‘forces’ and as such their existence radiates out to infinity.

          1.  Since events in one frame of reference cannot be compared with events in others does not rule out a common ‘Now’.

            No, but relativity says that it would be impossible in principle to find any empirical evidence for a single unique “Now”–either relativity is wrong about the laws of physics being identical in different inertial frames, or the notion of a “common Now” must be a purely metaphysical one that believers take on faith. Do you disagree that these are the only two options? If not, are you OK with the idea that the second might be true, and that you can have nothing other than faith to support your view on time?

            A mistake, which has a just little reference to our discussion, is the assumption of a ‘time line’  shortly after the ‘Big bang’. Fractions of seconds are quoted in the instants when ‘time’ would have been impossible to assess, both practically and theoretically,  since the measurement of time is relative to the adjacent mass.

            “Impossible to assess” by who? Since there were presumably no sentient beings back then, other quantities like mass and energy and spatial distance would have gone un-assessed too, but this doesn’t seem like a very good argument for their physical nonreality. And I’m not sure what you mean by “the measurement of time is relative to the adjacent mass”–in relativity “relative to” is always used in the context of some quantity (like speed) whose value is defined relative to the coordinate system you’re using (like the different inertial reference frames used by observers moving at constant velocity away from gravity). Are you talking about two different coordinate systems here, one used by an observer standing on a massive object and one used by a distant observer in space or something? If you are, note that proper time is not defined relative to any coordinate system but is a coordinate-invariant quantity, and real physical behaviors, like the half-life for a given particle to decay, are defined in terms of proper time rather than time in given reference frame.

        2. I’m sorry, but there is no reply button to your latest posting. Here will have to do.
          Since you cannot understand my postings I trust you understand I cannot understand yours.
          Time Line:-  The ONLY concept of time we have is the rate at which ‘time’ flows. This we measure by comparison with physical devices, clocks, oscillations, wavelengths. All of these are governed by the proximity to Mass. 
          SO, at the big bang no one can possible conjecture what time scale was involved.
          Simultaneity:- If I perform an action in one room and a friend performs an identical action in another room, one problem is the ‘trigger’ which synchronised the two events also, since the rate at which time ‘flows’ varies from one location to another, time dilation dictates the two events would not continue in synchronism. Similarly, if a third party performed the same action on the other side of the world, then the speed of light limits the accuracy of initial synchronisation. Taking the action several light years away would clearly make any notion of initial synchronisation impossible. HOWEVER, this does not mean that all actions, wherever performed, could NOT initially occur simultaneously anyway, whatever the rate at which the actions proceeded. The NOW of time.
          Schroedinger’s  Cat experiment is erroneously thought to confirm his views on the uncertainty of events, but has been wrongly translated. Even he was sceptical of deductions made in quantum theory. I think it has been forgotten that he proposed that matter was standing waves, in which case precise location of place and time allocated to any ‘particle’ would be impossible. 
          Similarly, Maxwell suspected ‘particles’ were vortices in the Aether. An idea he eventually disregarded.
          I’m not expecting you to agree with me, but I don’t propose to continue with this posting. When the Higgs field is proved to be a non-starter, perhaps we can continue with the Wave Theory of Matter. This will eventually explain how photons are transmitted via matter waves, hence the dependence of the speed of light on adjacent mass, and why Gravity is simply the refraction of the Matter Wave as predicted by Einstein as Space Time bending or warping.

      2. Time, space, matter and energy only exist as descriptors of each other.

        But there’s actually stuff between galaxies. Like photons. Or we wouldn’t be able to see them.

        1. Yes, within our present concepts of reality I would not be able or willing to dispute this, since we rely heavily (perhaps too heavily) on mathematics to express our interpretations.
          However, Dirac deduced that an high energy Photon (EM wave) produced two particles. The Electron and the Positron.
          Matter and anti-matter. Perhaps we can explain dark energy, and hence dark matter, if we consider how electromagnetic waves can combine and then dissociate  randomly in ’empty’ space? I believe this random ‘in’ and ‘out’ of existence is fully within the quantum theories.

  9. My sister, who is 19 years younger than me (same ‘rents) feels like she is time travelling whenever she talks to me.  

    I also believe she perceives some sort of time-dilation,  but she is too polite to mention it. 

    Picture attached to this comment is of me – from my high-school yearbook. I helped found the Chess Club. The pic is of me hand-crafting chess pieces, something I was well-known for back in the day.

  10. I’m going with, the human body will never travel forward or backwards in time… perhaps technology/metals/machines – could hypothetically(?), but not the human body. 

    There is evidence of time travel already, and that is our biological past and the path we see in our genetic history in the double-helix through evolution, and by or in our philosophical ‘form’ – but beyond that? Yeah, not really thinking it’s possible.  We are HERE now, we will never be THERE, then.

    Ya got one chance to live in this philosophical ‘form’, best to live it while you can!

  11. If incoming time travel were possible, there would (at the very least) have been many more recorded attempts on Hitler’s life amid strange circumstances.

    1. Where has it gone?
      The Quechua and Aymara see the past in front of them and the future behind them.

  12. how can we possibly have time travel in the next decade when i can’t even get a consistent 3G signal for my iphone?

  13. Time travel,as depicted by popular fiction, is not possible. What’s he talking about is that in a rotating universe, you can go into someone else’s past. For example, the Andromeda galaxy is 2 million light years away. If our universe is rotating, you visit it in its past somewhere between 2 million years ago and present day. But when you get back to good old Earth, you will still be in Earth’s future, not its past.

    1. Not true, in a rotating Godel universe, spacetime is twisted in such a way that you could actually go on an extended journey that would take you back to Earth at a point in the past light cone of the moment you departed (on the simplest type of closed timelike curve you would return to Earth at the exact moment you departed, but the path can be distorted a bit so you arrive on Earth earlier and then wait around until the moment of departure).  See the time traveling and general relativity section of a page by a physicist who wrote his Ph.D. thesis dealing with rotating universes, which shows a light-cone diagram and comments “A material object can travel from an event A, around the universe, and then back to event A. The curve can even be modified so that the object arrives before it was launched. If it then waits to see the launching, the timelike curve is closed.” Godel himself said something similar, as quoted on p. 86 of The New Time Travelers: “the light signal will come back at exactly the same moment at which it is sent … Since a light path can be approximated as closely as you wish by a path of a material particle, you can even travel into the past on a rocket ship of sufficiently high velocity.” Of course, our universe probably isn’t really rotating, at least not at a rate sufficient to make this possible, but such a universe is an allowed solution to the equations of Einstein’s theory of general relativity. And there are other theoretical solutions in general relativity that allow for backwards time travel without involving the whole universe, like traversable wormholes.

      1.  There are a number of assumptions of that post that are wrong. First, the Mach principle is valid since distance masses influence what happens locally by gravity; that’s what general relativity is all about: gravity. Second, it assumes that line cones are always perfect. True when you are dealing with only constant velocities but when you throw acceleration into the mix, they’re only perfect cones at the observer. When you move away from the observer, they become distorted. They can even cross themselves. You cannot use them in general relativity because they do not account for the variable bending of space.

        But I’m not surprised by this. Physicists have a hard time believing that general relativity causes the expansion of the universe. Anything more complex than that is bound to be full of errors.

        1. First, the Mach principle is valid since distance masses influence what happens locally by gravity; that’s what general relativity is all about: gravity.

          Your notion of Mach’s principle is far too vague, it isn’t just about gravity “influencing what happens locally” (if it were, then Newtonian gravity would satisfy Mach’s principle, but it clearly doesn’t), it’s about all acceleration-based effects (such as the centrifugal force felt in a rotating frame of reference) existing solely because of acceleration relative to distant masses. General relativity says that if you had a universe which was totally devoid of matter save for a single bucket filled with water, the water would remain flat if the bucket weren’t rotating, but would bulge out if it were. So, most physicists say that general relativity does not perfectly satisfy Mach’s principle.

          Second, it assumes that line cones are always perfect. True when you are dealing with only constant velocities but when you throw acceleration into the mix, they’re only perfect cones at the observer. When you move away from the observer, they become distorted.

          The past light cone of an event E is a perfectly well-defined notion (it’s the set of all points in spacetime that could send a signal moving at the speed of light or slower that would reach E) regardless of its shape, there is no assumption in general relativity (or on the page I linked to) that light cones would be “perfect” in the sense of having nice straight sides (which I assume is what you mean by “perfect”). Perhaps you are referring to the diagram on the page I linked to that illustrates light cones with straight sides, but it’s just a schematic intended to show the angle of the light cone in the immediate vicinity of the point on the tip, a common convention in general relativity diagrams (seen in many diagrams in the most famous GR textbook, “Gravitation” by Misner/Thorne/Wheeler).

          You cannot use them in general relativity because they do not account for the variable bending of space.

          Again, nothing in the definition of a “light cone” requires that the sides look like straight lines, and light cones are widely used in general relativity. Why would you make such confident statements dismissing something a physicist wrote about an area of physics you clearly haven’t made any technical study of? Beware the unskilled and unaware trap of overconfidence in your own judgments about a subject you are fairly ignorant of…

           Physicists have a hard time believing that general relativity causes the expansion of the universe. Anything more complex than that is bound to be full of errors. 

          Huh? Physicists today have no particular problem believing in general relativity’s prediction of an expanding universe. But in any case, that’s a real-world question about whether the actual universe matches up with the predictions of the theory. The question of whether general relativity theoretically allows for closed timelike curves is a purely mathematical one that doesn’t depend on whether you believe the theory is actually correct in the real world. (as I said, no one believes that our universe is a rotating one!) Godel discovered the possibility of closed timelike curves in GR in 1949, and the calculations have not been challenged by any mainstream physicists or mathematicians for all this time, are you seriously suggesting that all of them could be making a mathematical error in their calculations of what is allowed theoretically in general relativity, and that you are qualified to judge them all wrong even though you have no detailed technical understanding of the subject?

  14. Here is a link to THNKR’s follow-up episode, which goes in-depth into Dr. Mallett’s personal theories on the potential of time travel and time machines. On Thursday, he will answer questions from the cybersphere, and on Friday, he will discuss five things he’s currently really excited about. Check back to see the whole series!

Comments are closed.