In this article I want to explore the idea of using time travel to affect the course of history, to right wrongs, and otherwise impose one’s own wants and wishes onto the past in order to shape the course of historical events. Righting the wrongs of the past is a common theme in time travel fiction, and it represents an attractive proposition to us: Can we save the unsavable? Supposing we had access to a classical time machine,1 could we do it? My answer is… Well, probably not.
I don’t want to get into the muddy issue of whether time travel is theoretically possible or not – I don’t know the answer to that.2 Physics literature has not yet conclusively ruled it out,3 though the literature doesn’t often concern itself with what we might think of as classical time machines, but rather those of a distinctly different and potentially more plausible variety.4 I’m not knowledgeable on the detailed physics of the subject, so I’ll avoid entering into a too scientifically technical discussion.
One factor determining the efficacy of time travel for the purposes of rewriting history is the characteristics of the kind of time we’re dealing with – by which I mean the nature of time itself. It had long been assumed that time was a linear progression – an interminable series of moments, cascading one after the other at a set speed of motion that was the same for everyone. With Einstein’s discovery of relativity, we now know that to be incorrect. Time passes at different rates depending on one’s position relative to gravitational bodies. Though relativity gives us a glimpse into the potential mechanics of time travel, it also gives us some stumbling blocks.
The basic characteristic of travelling back in time is that the traveller must reach his destination before he starts off. If he accomplishes this, we may say that he has travelled back in time. So, how might he do that? The tantalising solution is for him to travel so quickly that he arrives before he leaves. There is something intuitively objectionable to the idea that he could do this, but before the discovery of relativity it was not known what exactly the limitation was. It would seem to run contrary to the chain of causality – an effect (his arrival) would be occurring before its cause (his departure). But the questions of how fast he could really go, and why he could go that fast and no faster, were yet undetermined. With relativity, we know the answer to be the speed of light.
The speed of light is the ceiling that no body of mass can ever reach (so far as we know), because the amount of energy required to propel it ever faster increases exponentially as the object continues to accelerate. The point at which a particle with mass is travelling at the speed of light, its mass will be infinitely large, the energy required for it to further accelerate will be infinitely taxing, and time as experienced by that mass will have stopped. And still it won’t have reached its destination before it set off. Though we know a modicum about time in this relativistic sense, we still have little understanding of the consequences that, should a person have travelled back in time, might present themselves.
One type of time popular in fiction is (what I will call) the timelines theory. It is similar to the linear model of time progression that I mentioned earlier in that time exists as a series of events, with each event definitively dependent on an event that happened immediately prior. It attempts to resolve the question of what happens when a time traveller moves back in time and changes the outcome of an event. It is, in my opinion, the most coherent model for resolving the common time-travel paradoxes.
In the timelines model, the traveller can be said to have created a fork, allowing two simultaneous timelines to progress in tandem. By moving backwards in time from his current time T1, he has created a brand new event at destination T2 on his timeline X. However, his timeline X already has an event at his destination time T2 that led to the series of events that enabled him to travel back in time in the first place at T1. If T2 is overwritten with the new event (the time traveller’s arrival), that could interfere with the series of events that led to T1 (the time traveller’s decision to go back in time). If, on the other hand, T2 is not overwritten, then the traveller has not travelled back in time.
The fork in timelines resolves this contradiction, and allows there to now be two timelines. The original timeline – timeline X – continues as normal, meanwhile timeline Y pops into existence as a consequence of there being two discrete events at T2. One timeline has the time traveller in it (in his present state), the other does not. He has moved from timeline X to timeline Y (into the past relative to T1 in timeline X).
One of the implications of this model is that there is no obvious method for the traveller to return back to timeline X. If he travels forwards in time, he will be travelling forwards along timeline Y. If he travels backwards – back to T2 – he will simply create another event and a consequent timeline Z. He is, effectively, stranded. If he was hoping to affect the lives of the people of timeline X, he would have failed. He may have been able to affect the lives of the people of timeline Y, but these people would not have existed had his time travelling not caused them to come into existence. He has not been able to legitimately save anyone.
Another implication of this model is that it would suggest there to be a multiverse (one universe for each timeline). This model would appear to be fully compatible with the popular conception of a multiverse, where there is a universe for every possible scenario.5 In this circumstance, having a “timelines model” would be somewhat redundant, as every possible scenario would have its own universe (and so its own timeline). There would not just be timelines as a result of time travel, but for every quantum event that occurs in any of the universes at every conceivable moment.
The immediate alternative to creating a fork is that the traveller, when arriving at T2, does in fact overwrite the previous event. This would effectively erase all the events that occurred between T2 and T1 (from the traveller’s present to his destination in the past). The traveller will have made it to T2 and will have set in motion a new course of events. This is the traditional model of time travel that leads to the famous paradoxes that mess with causality.6 In the timelines model, there are no problems because there are no unresolved questions of causal dependency – one timeline allowed the traveller to go back in time, another timeline sees the consequences of his time travelling.
So, supposing this is the model we have to work with, should a person try to rewrite history for the better? Well, it depends. The problem, as I see it, is about erasing events that have already happened – what does it mean to ‘undo’ an event? By travelling back in time, you are not really erasing it. Somewhere, somehow, the people you want to save already suffered what you wish to save them from, and nothing you do can alter that. Indeed, you yourself are proof of it, because this was your very motivation to travel back in time in the first place, and so it has left its mark on you just as it did them. In effect, you are deleting everyone from existence at T1 – not genuinely undoing what happened in the events that led up to it.
However, we have yet to factor in a consideration of the events that might occur post-T1. If more bad is going to happen, then you may well see it ethically preferable to halt the continuation of that timeline and create a new, better one. But it will be a tough decision, because overwriting a timeline is akin to murdering everyone in the universe – at least in the sense that you are deleting that universe and everything in it. Whether you think that’s bad or not of course depends on your derivation of ethics.
So, under this model it might be advantageous to rewrite history in some circumstances, but likely only to be done in truly exceptional circumstances. The common thought about using time travel to go back in time to assassinate Hitler, while it has good intentions, falls into problems with this model. The Holocaust will still have happened, and you would only be erasing any remaining traces of its effects. Unless the present day is in some way unacceptable to you, and you think the assassination of Hitler would improve it a great deal, it would not make sense to attempt it.
Another way to think about it would be to consider what would happen if you went back in time to give yourself advice. You are not truly giving advice to yourself, for you have grown up without that advice and because of that fact decided to travel back in time to redress the situation. Because you are not giving advice truly to yourself, you must be giving it to someone else who shares your physical make-up. This person will then grow up with that advice and live a different life, but you yourself will always be you – a product of not having received that advice. You will not have undone your situation. This is assuming, of course, that your action of meeting your previous self does not interfere with his future decision to travel back in time, causing you to vanish from existence in a cloud of conundrum and mystery.
There is another popular understanding of time travel – that it can create interdependencies between events. This can lead to interesting consequences, such as looping. Consider the scenario where a person sees a photograph of his older self in the past. This inspires the person to investigate time travel and devise a method to travel back in time. During the course of his travels, he has his photograph taken, and it winds up in the hands of his younger self, beginning the causal loop all over again. This is what I like to call the ‘whatever happened, happened’ model (after Lost).
The problem with this view is that it seems, on the surface at least, to be internally inconsistent. How does the loop begin? The answer as far as I see it is that it doesn’t ‘begin’. One of the implications of this model is that events are preset and determined. Time is interconnected, such that the system effectively operates as a singularity. There is no single ‘present’ because everything has already unfolded, past and future. Potentially, every moment in the temporal singularity is occurring simultaneously (if we consider it from an external reference point).
Such a universe would preclude the existence of free will (at least insofar as having any power to affect the loop itself) – each person would be locked into a set destiny, or else (if not somehow otherwise limited) this would eventually undermine the structural interdeterminacy of the system. It would, for example, be a person’s destiny to travel back in time and fulfil the conditions of the time loop, or else there would be no time loop and no interdeterminacy. In order for temporal recursion of events, a past event must depend on a future event, and the future event depend on the past event. These events would be both the antecedent and consequent to each other, and must be fixed – otherwise free agents operating in the system would be able to affect these events, dislodging the loop.
Thinking about these matters, the problem of determining which kind of universe we are in (and I expect there will be other varieties I haven’t thought of) seems on the face of it insoluble. From the perspective of the time traveller – who is the only person seemingly able to gather evidence on the subject – there are no signs pointing to which kind of universe he may be in. If he is in the timelines model, he has no way of finding his way back to timeline X even if he did find the answer. In fact, no traveller would be able to travel from a timeline without this knowledge to one with that knowledge (assuming the knowledge wasn’t forgotten or misplaced), because he would only have access to the main trunk of his own timeline branch.
In the overwrite model, the timeline between the traveller’s departure and his destination would disappear, but the traveller would have no means of knowing about it since he would be creating a new one in its stead, which would be identical to timeline Y had he created a fork. As for the ‘whatever happened, happened’ model, the traveller may be aware of his fate or he may not – but in any case, he would have no proof that his loop would continue indefinitely and was a true instance of temporal interdeterminacy.
My personal intuition on the plausibility of these conjectures is that backwards time travel is impossible. It seems to me to be too much to expect the universe to have a mechanism to split off into two every time a causal anomaly occurs – I would find it more likely if, if such anomalies are possible, that the universe would simply break. There is no guarantee that the universe is stable.7 There would also seem to be no greater barrier to be overcome than that posed by the speed of light, which seems as much of an intractable problem to me as any I can think of. If it is possible, the timelines model would at least go some measure towards explaining why we haven’t met any time travellers yet.
Time travel is an exceptionally mysterious and interesting subject to me. While on the surface it hints at the idea that nothing is permanent, a deeper reflection shows that there is no meaningful way to annul what has been previously committed. In one way or another, what’s done is done, and there may be a philosophical satisfaction to be gained in accepting this. So, let’s not take the past too seriously.
The technical term for what the layman would consider a ‘time machine’ is a ‘Wellsian’ time machine – where a brave adventurer straps himself in, flips a switch and catapults himself forwards or backwards in time. It’s named after H.G. Wells, author of the sci-fi classic The Time Machine. ↩
Well, that’s not strictly true: I do know to a certain extent. The problem we have with time travel as it currently stands is that we can only go in one direction. Given a relativistic universe, it’s possible to travel far into the future by moving very quickly relative to another body of mass, such that the amount of time you experience is less than that of the observers on the body of mass. This has been proven experimentally by using atomic clocks and planes in the Hafele-Keating experiment. ↩
On the subject of the theoretical plausibility of time machines, the Stanford Encyclopedia of Philosophy writes:
The physics literature contains various no-go theorems for time machines, i.e., theorems which purport to establish that, under physically plausible assumptions, the operation of a time machine is impossible. We conclude that for the time being there exists no conclusive no-go theorem against time machines.
Physics literature, in its discussion of time machines, often talks regarding Thornian time machines (named after Kip Thorne – see Morris and Thorne 1988; Morris, Thorne, and Yurtsever 1988) – devices hypothetically possible to introduce CTCs (closed timelike curves), such as those represented by wormholes. See the Stanford Encyclopedia of Philosophy page for more details on these devices. ↩
In quantum mechanics, this idea exists as the ‘many worlds’ theory – an alternative to the more popular Copenhagen interpretation. Instead of there being superpositions of states (where a quantum event can be said to have both occurred and not to have occurred, before its observation forces the universe to ‘make up its mind’), the chain of causality splits, and there are now two worlds – one where the event occurred, and one where it did not. ↩
The most famous example is the grandfather paradox: What happens if you travel back in time and kill your own grandfather before he meets your grandmother? Do you disappear? Does the universe break? Is it even possible? ↩
Of course, from any perspective that may consider it, the universe will always be stable: If it disintegrated there would be no perspectives left to consider its instability. In this way, the anthropic principle guarantees us to be in a non-destructed universe. ↩