Is Time Travel Possible?
People are always travelling in time. The average person will be born, travel through somewhere around 70 years and then die; depending on how we define a person they'll either then end their journey, or continuing travelling (in an ever more random and spread out pattern) till the end of time. Presumably, therefore, when people ask whether 'Time Travel' is possible they are not referring to the usual tendency for people not to exist only instantaneously, but to be stretched over a period of time, in much the way they stretch along a certain distance in the three spatial dimensions - and are instead asking whether it is possible for people to travel through time at a non-standard rate , or even instantaneously - disappearing from one time and rematerialising straight away (from their perspective) at another. I will try to show in this essay that 'time travel into the future' is not just possible, but is simply what we observe from day to day and that if we are to have a serious discussion about whether time travel is possible it must be about travel into the past.
It might be argued that whilst forwards time travel normally occurs, it is not necessarily the case that we should be able to travel faster into the future than normal, or even than other people around us. This, however, does appear to be possible, and in fact, as Einstein suggested, we need only travel at a fast speed to reach the future with a shorter amount of time passing for those moving quickly that those that are stationary - and such a principle has even been shown to work with very accurate clocks on board aircraft and spacecraft that we already have the technology to build. There appears to be no physical limit to how fast we might pass through time - the only upper limit on speed being that of the speed of light, at which point time will appear to stop completely for the person moving. Yet we do not even need to be constructing very powerful vehicles to travel quickly through time - this is the very point of relativity - since an astronaut is no more travelling through time that someone on Earth, who, to an observer in a different galaxy, is already moving at a very fast speed, and travelling at a non-standard rate through time - it is all a matter of perspective.
As to the possibility of instantaneous travel, whilst this might technically be different from travel through time that is so quick as to appear instantaneous (yet supposedly still takes at least 1 plank-time, the theoretical minimum amount of time according to quantum theory), I cannot see how this would make a significant difference, and why any of the philosophical objections to very quick time travel would not hold for instantaneous travel. I will not, therefore, discuss this separately, so long as there appear to be no restrictions as to how fast travel might be, other than that it might have to take a matter of milliseconds to occur.
If we are going to determine exactly what we mean by time travel, as opposed to simply sitting around and watching time pass, we must first define exactly what we mean by time itself, and this turns out not to be a simple task at all, especially if we try to formulate time with a direction. If we are going to agree that forwards time travel is possible we will have to show that in some real way time looked at one way is different to it looked at another way - i.e. that it doesn't just matter that you get the events in some pattern, or even the right pattern, but also that you get the first one at the beginning and the last one at the end - time must not be just a string of events that can be taken either way round.
Various suggestions have been made as to what it consists in to say time is moving forwards. As Lewis points out, physicists have particularly settled on three possibilities - that the universe is expanding through time, that entropy (disorganisation) increases through time, and that radiation expands outwards through time. All of these seem clearly to be empirical observations that have been used to characterise some things that occur as time goes forward, but don't seem to be satisfactory answers for what time going forward consists in. Perhaps some might argue that we actually perceive the world in the way we do because the universe is expanding or because it is getting less organised, but I don't think they mean that we would actually get back to times we had previously existed in and be able to see the past if the universe began to collapse or if order (even temporarily) increased. Even if we perceived things differently - with other galaxies getting closer, or broken eggs jumping back into their shells, this would not be the same as moving in the other direction along the time line, ending up at a time which we had characterised as occurring 'beforehand'. Indeed if we are to take one of these definitions to be true, then by the very definition we wouldn't see things differently, because even if the universe were collapsing we would still see snapshots of it further and further apart, even if it were claimed that we were seeing this backwards. All-in-all a very strange world we would live in if one of these definitions held.
Mellor considers the example of backwards clocks, and asks what a clock would look like when time was going backwards. More specifically, he points out that it would be difficult to tell the difference between time going backwards and a clock turning anti-clockwise - and indeed the only way to tell is presumably that various other things would still seem to be going 'forwards'. What, then, would it mean for time to be going backwards, since presumably we would go backwards with it? The only reason that we are able to tell that a video rewinding is not in fact fast-forwarding is because we are going forwards through time, and thus we don't see everything else going in reverse, and also that we see new things happening, rather than observing the events of the last day playing back to us in reverse. So if time is to have a direction where walking backwards round a room does not count as time travel there must be some way of distinguishing between the bits of time we label the future and those that we label the past which does not simply rely on the standard order in which things occur, or the direction in which things normally travel.
One way of formulating such a distinction centres on this idea of 'new' and 'old' events - namely that we remember things that have happened in the past, and are not yet aware of things that are going to happen in the future. Such a theory based on memory as the measure of direction of time is certainly attractive for the fact that it makes it a very easy thing to determine whether something is in the past or future. One suggested problem with such a definition, is that if we do not refine it we will end up with an idea of the direction of time totally incompatible with normal ideas about time-travellers, since presumably if a time-traveller remembered the Roman era occurring after 2000 A.D. (because they had travelled back to then) we would be forced to say that it in fact occurred after. This attack seems somewhat unfair - since we made no claims about the order in which people remember stuff (which could clearly be misleading, and people do get confused about the order in which things occurred), and we can say simply that the things which most people remember happened in the past, since, as long as we are trying to claim that there is a normal direction of travel we can use the vast majority who are not time-travellers to be the arbiters of time's direction.
Overall, therefore, it would seem to be the case that it is travel into the past that will be of interest to us, and it is the objection to such a possibility that we must consider - and I hope to show that none of these are satisfactory to rule out the possibility completely. The first objection must be to ask whether there are simply physical constraints on time-travel, since it being a contingent fact that the way the world is aligned simply will not allow it will be an easier thing to prove than that it is a necessary fact. It might be argued that this is a far less interesting and important question than whether there is actually a logical constraint on time-travel, and that it is in fact necessarily impossible, yet, given that we do live in this particular universe, I do think that it is an useful question to consider.
The most common way of arguing that it will never be possible for humans to travel in time is simply to pose the question that if it were ever possible why we haven't seen any time-travellers coming back from the future. I would argue that in fact this isn't a particularly relevant question to ask, and it firstly will never go so far as to prove anything - perhaps time-travel will be possible but future generations, realising it is a bad idea, will decide against it - and regardless will only ever serve to suggest that time-travel to a time before the invention of time machines is impossible. In fact, many people do believe this less strong claim, including Stephen Hawking, who argues that it would not be possible to travel to a time before the invention of time machines. If we posit a time machine that consisted of a special room, into which we would step, and through which we could be transported to the same room (or perhaps with the more advanced version any number of like rooms in other places) at a different time, we obviously would have to wait till such a room had been constructed before we would be able to travel back to it. Likewise with another suggested possibility for a time machine, which would involve taking two blackholes close to each other, pulling them apart (with some cutting edge tractor) and using the immense gravity well for instantaneous travel between - which in terms of relativity means time travel - which could not get you back to a time before you build this gateway.
So if a large number of well respected physicists are not willing to rule out the possibility of time travel on the basis of physical features of our universe it is probably unwise for us to dismiss it on these terms, and we should, instead, turn to more philosophical questions of whether it would be logically impossible for time-travel to occur, and whether it would cause contradictions. Luckily for us, there has been a large amount of debate as to whether such contradictions will arise, and many believe that they will.
Mellor highlights the problem that might be created by 'causal loops', when we have an event P that causes another event, Q, which due to time travel causes P. If such loops occurred we'd appear to have a very strange series of events with which, whilst consistent and where it could be traced which led to which, we could never explain how the entire loop came about in the first place. It is not, however, clear why such loops should in fact be a problem. If they are consistent there does not seem to be a reason why they shouldn't in fact spontaneously appear on the timeline, and it could well be that given the technology for time-travel such events could arise. It might even seem, if we are to follow some sort of chaos theory, that these would have to occur, since every tiny movement of a time-traveller in the past would alter the whole course of history, and without such a movement the time machine might never have been invented.
Furthermore, there is one causal loop that has long been considered and discussed as to whether it is possible - namely the whole course of the history of the universe. Various scientists have suggested that time may be circular, with the big bang caused by the universe slowing down in its expansion, and eventually collapsing in on itself in a 'big crunch' - such a momentous event would surely provide the enormous amount of energy needed to cause the big bang. Regardless of whether this is the case or not the example of the whole of time illustrates something well - that whilst we may explain causation in terms of every event in the chain causing the next, we seem never to be able to explain the chain as a whole. People seem unable to attempt to explain what caused the earliest event (some would say the big bang) because that raises the question of what caused the cause. So we could have a causal loop where every event within the chain would explain the next, but it would not be possible (and perhaps not even sensible) to attempt to explain what caused the whole loop.
Another problem for the causal loop, however, as Mellor debates is that if we have a causal loop it would seem to be the case that any event could be said to cause itself, since if A causes B and B causes C and C causes D and D causes A we must surely be able to say that A causes C, D, and A. This relies on a mistaken presumption that causation is obviously transitive, which is not necessarily the case by any means. Many formulations of the concept of causation deny that it can be transitive, especially the counter-factual theory, which I think would be a particular one to look out for when debating time-travel - so it may be possible to argue that each of these things cause the next, but it does not follow that the first causes the last.
There is a complementary problem to that of causal loops, so similar that it might be considered an anti-causal loop (or perhaps even a causal-mobius-strip) - namely that created by the Grandfather Paradox. Put simply, the Grandfather Paradox highlights the problems that would occur if a time-traveller were to go back to the past and kill his grandfather before his father was born; if this happened his grandfather would never have his son, who in turn would never have his son who would therefore not grow up to become a time-traveller, and thus not kill his grandfather, who would live to have children. This would cause greater problems than the causal loop, because rather than being self-consistent but not explainable in terms of how the whole thing arose, we can explain how it came about but it is inconsistent. Lewis suggests that it even disobeys the law of the excluded middle, since the grandfather would be both dead and not-dead.
Lewis' solution to this problem is one that I would entirely agree with - namely that whilst it would indeed cause enormous problems if such a paradox ever did arise, people would simply fail to kill their grandparents before their parents were born if they ever tried. I would suggest that so destructive would a time-paradox be that it would cause the whole timeline to collapse - not just after the event (which would seem a fairly ridiculous proposal) but at all points along it - the timeline would cease to exist or ever have existed, and time would never have been. Quite clearly, however, this isn't the case - time hasn't been destroyed, and exists (at least now) and so we can simply assert that no time paradox will ever occur, even if one contingently could. The reason that people cannot go back and change the past is simply that they would only end up changing the past to how it is now - i.e. how is always was and always will be. As Lewis puts it 'The events of a past moment could no more change than number could' - it seems a bizarre sort of fatalism that doesn't just deny that the future is not yet set (though I would argue it is) but also that the past is not.
To put it another way, it may be true that the indexical sentence 'The Queen of England now is Elizabeth II' only for a particular stretch of time, but the non-indexical sentence 'The Queen of England when this essay is being written is Elizabeth II' will not change, but will be true forever. Likewise I don't have to go back in time for the statement 'The monarch of England 200 years before this essay was written was not Elizabeth II' to be true - it is true now, and cannot be changed. So we cannot change these non-indexical facts about the past, and since one of those is that my grandfather lived to have a child I couldn't go back in time and kill him before he does.
Another way to look at the issue might be to consider how we would view things if we discovered that such an event had occurred. If, for example, I found out that my 'grandfather' had not lived to the age he must have for my father to be conceived, what I would say would be 'Oh, he must not have been my grandfather then'. In the Terminator films a robot comes back from the future to try to prevent a woman having a baby who will grow up to be the leader of the rebel uprising against the robots, and another is sent when he is growing up to stop him doing so. It emerges that the robots were created by reverse engineering a chip that came from one of the ones that had been sent back; yet by the end of the two films all the chips from these machines were destroyed. Do we, as an audience, think 'But that's created a paradox, since the robots would never be created then!' - I would argue that we do not, instead we simply say 'Oh, there must have been another one', and make comments about smelling a sequel. I would suggest that such pedantry is totally justified, and we feel unsatisfied when such loose ends are created and appear to go unnoticed by the writers - because we have a good instinct that if a paradox appears to have been created, it's simply that there's something that has been missed, and something is not how it seems. I would argue that this will always be the case, that we'd simply discover that our 'grandfather' wasn't really our grandfather if we killed him (perhaps we were adopted) - and that, overall, there are no problems with the possibility of time-travel so robust as to rule it out altogether.