Author: Taylor W. Cyr
Word Count: 1000
Time travel is familiar from science fiction and is interesting to philosophers because of the metaphysical issues it raises: the nature of time, causation, personal identity, and freedom, among others.
It’s widely accepted that time travel to the future is possible, but the possibility of backward time travel remains hotly debated. This article will sketch some models of backward time travel (hereafter simply “time travel”) before addressing the main objections to its possibility.
1. Models of Time Travel
According to the standard model of time travel, time is linear, so a time traveler’s journey may be depicted along a single timeline, with some events that occur earlier in the timeline’s being experienced as later by the traveler:
On another model, time travel results in the creation of a new universe that branches out from the same trunk (shared past) as the original:
A third model of time travel maintains that there is a second temporal dimension, and so, in addition to times, there are “hyper-times.” On this model, time is more like a plane than like a line, and a time traveler may, in returning to an earlier time, reach that time at a later hyper-time, with the result that the aforementioned time bears different properties at the different hyper-times:
2. Changing the Past
It is natural to suppose that time travel would change the past, which many believe is impossible. Changing the past would require that the past have a certain property at one “time” and then lack that property at another “time.” This is incoherent on the standard model of time travel, which maintains that time is linear (there is no “second time around”), so the standard model precludes changing the past.
But time travel doesn’t require changing the past. We may distinguish changing the past from affecting the past, where the latter requires only that the time traveler’s travels have effects in the past. For example, suppose a time traveler finds her younger self and attempts to convince herself not to time travel. Assuming the standard model of time travel, she will fail to prevent herself from time traveling, but the attempt will affect how the past was “all along,” so to speak. From the outside, the scene will look like an ordinary conversation between two people, but, assuming the time traveler remembers the scene, she will remember and older version of herself trying to convince her not to time travel.
Moreover, according to the other two models of time travel, one and the same time may exist in two different universes or hyper-times, and so it isn’t obviously incoherent to state that some past time may have a property at one “time” (either in one universe, or at one hyper-time) that it lacks at another “time” (in another universe, or at another hyper-time).
3. Causal Loops
Consider some events from the television show Lost. At one point, Richard gives a compass to Locke, telling him to return it the next time they meet. Locke then travels back in time, sees a younger Richard, and returns the compass, which Richard keeps until he gives it to Locke in the aforementioned meeting.
The Lost compass is strange. It was not created in the usual way—in fact, it has no creator! It appeared (with Locke) at time t1 (when it was given to Richard), remained with Richard at a later time t2, and then was given to Locke at t3, when Locke set out for t1, resulting in a “causal loop.” At each time t1-t3, there is a causal explanation for the compass’s presence by reference to the prior stage in the loop. But no explanation can be given for the loop itself. (Where did the compass come from to begin with? There is no answer.)
Now, if such cases are impossible, this might cast doubt on the possibility of time travel. As David Lewis says in response, however, such cases “are not too different from inexplicabilities we are already inured to” such as “God, or the Big Bang, or the decay of a tritium atom,” all of which are “uncaused and inexplicable” (1976: 149).
Note that this objection assumes the standard model of time travel, since these strange loops do not necessarily result from time travel on the other models. Moreover, it may be possible for there to be cases of time travel that don’t generate causal loops even assuming the standard model.
4. Time Travelers’ Abilities
Suppose Tim time travels and attempts to kill his Grandfather before his parents are conceived. Assuming Tim has a gun, is a good shot, etc., it would seem that Tim can kill Grandfather. But Tim can’t kill Grandfather, for doing so would preclude his own existence. Tim both can and can’t kill Grandfather: that’s a contradiction, so we should give up the assumption that led to it, namely that time travel is possible.
This is the Grandfather Paradox, and it is the main objection to the possibility of time travel. Here are two responses, both of which assume the standard model of time travel.
First, one might understand “can” claims like “Tim can kill Grandfather” as claims about what is possible in view of certain facts—and which facts are held fixed is determined by the context of utterance. For example, in view of Tim’s possession of a gun, his reliable aim, etc., it is true that Tim can kill Grandfather. But if we also hold fixed the fact that Grandfather lives, then Tim’s killing Grandfather isn’t possible, and thus he can’t kill Grandfather. So, there is no contradiction; it is true that Tim can kill Grandfather holding certain facts fixed, and it is false holding more fixed, but the claim is not both true and false in the same context.
A second approach denies that Tim can kill Grandfather. This denial follows from certain independently motivated views of agents’ abilities, and it avoids the Paradox by restricting the freedom of time travelers.
Perhaps time travel is (metaphysically) possible, but it doesn’t follow that it’s technologically feasible, or that it will ever actually occur. Only time will tell.
 While not the first philosophical discussion of time travel, David Lewis’s classic 1976 essay “The Paradoxes of Time Travel” popularized the subject in metaphysics. For a recent philosophical discussion of time travel—an excellent summary of several facets of the debate, as well as some new developments—see Wasserman (2018).
 By “possibility” I mean metaphysical possibility—consistency with the laws of metaphysics, such as the laws of causation, identity, etc. For more on the discussion of the various senses of possibility we might be asking about in connection with time travel, see Wasserman (2018, chapter 1), and see the rest of the same book for a summary of the debate about the metaphysical possibility of backward time travel.
 There are other objections, but there isn’t space to consider all of them here. One objection concerns its likelihood rather than its possibility. As we will see below, there are certain things that it would seem time travelers cannot do, and so if time travelers attempted the impossible, something would prevent them from succeeding (perhaps the time traveler would have a change of heart, or perhaps she would slip on a banana peel, or…). Horwich (1987) argues that since backward time travel would result in such improbable events, this casts doubt on the likelihood of time travel. See Smith (1997) for discussion and a response to Horwich.
 See the first figure. Reprinted from Wasserman (2018, chapter 3) with kind permission of Ryan Wasserman and Oxford University Press.
 For developments of the hyper-time model, see Meiland (1974), Goddu (2003), and van Inwagen (2010).
 If we graphed the two dimensions of time on a plane, with the temporal dimension along the x-axis and the hyper-temporal dimension along the y-axis, as in the third figure, time travel would amount to moving leftward (back in time) and upward (forward in hyper-time).
 As Brier explains, “One cannot change the past or undo what has been done. Rather, what is at issue is whether one can affect the past; that is, by a present action cause something to have happened which would not have happened otherwise” (1973: 361).
 For a simple example of this from science-fiction, see the film Interstellar. After leaving Earth, Cooper is able to send messages back in time, and he uses his first message to try to get his daughter to make him stay on Earth, as seen here.
 For another example of affecting (but not changing) the past, see J. K. Rowling’s Harry Potter and the Prisoner of Azkaban. An especially excellent case of time travel occurs toward the end of the book when Hermione takes Harry back in time, allowing him to save himself from dementors. In the film version, we see Harry attacked by (but saved from) dementors here, and then we see Hermione take Harry back in time here, and finally we see Harry save himself here.
 It is contentious whether these models of time travel really allow for changing the past. See Smith (1997, 2015) and Baron (2017) for arguments against, and see Law (Forthcoming) for a response.
 The first of these occurs in the third episode of season five, “Jughead,” from 39:44-41:19, and the second scene occurs in the first episode of season five, “Because You Left,” from 29:30-34:34.
 For example, suppose I travel back in time by twenty seconds but set my machine to a destination on the other side of the planet. Presumably my appearance in the past will not have any causal consequences across the globe, despite its occurring twenty seconds earlier than my departure, and thus no causal loop will be generated. For a similar example, see Hanley (2004: 130).
 On the other models, there is no reason to think that Tim can’t kill Grandfather, for doing so would preclude Tim’s future birth in the new timeline (the new branch or hyper-time), but Grandfather would not have been killed in the original, and thus Tim is still born in that timeline.
 See Kratzer (1977).
 While Lewis’s (1776: 149-152) influential response to the Paradox also relies on the Kratzer semantics for “can,” his proposed resolution is slightly different, for he sees the fact that Grandfather lives as one that it would be illegitimate to hold fixed. Holding it fixed, he thinks, amounts to “fatalist trickery,” as such a fact “is an irrelevant fact about the future masquerading as a relevant fact about the past” (1976: 151).
 See Vihvelin (1996).
Personal Identity by Chad Vance
Free Will and Free Choice by Jonah Nagashima
About the Author
Taylor W. Cyr is a lecturer of philosophy at Washington University in St. Louis. His main research interests lie at the intersection of ethics and metaphysics, including such topics as free will, moral responsibility, death, and time. His work has appeared in such journals as Ethics, Philosophical Studies, Philosophical Quarterly, and Erkenntnis. https://taylorwcyr.com/