Time Travel Movies

by Brian on August 29, 2003

I’m teaching a freshman seminar on time travel at Brown this year, so I’ve been watching a lot of time travel movies as ‘preparation’. I always knew that many time travel movies don’t make a lot of sense on a bit of reflection. What surprised me on recent re-watchings was that some seemed unintelligible even on relatively generous assumptions.

Philosophers normally break time-travel stories into two categories: those that do make sense within a ‘one-dimensional’ view of time and those that don’t.

The ones that make sense on a ‘one-dimensional’ view never have it the case that at a particular time something both is and isn’t the case. They don’t require that the direction of causation always goes from past to future, that would stop them from being time travel stories after all, but they require that there be a single complete and coherent story that can be told of the history of the world. Some philosophers are known to reserve the label ‘consistent’ for these stories, but that’s probably a bit harsh.[1]

Some stories keep to this constraint, even when they are under a lot of pressure to break it. The first Terminator does, the second Terminator might (though it’s normally interpreted as violating it), and both 12 Monkeys and it’s inspiration _La Jetée_ display quite a bit of ingenuity in telling an involved time-travel story that has a coherent one-dimensional history.

But obviously this kind of constraint is not a universal norm among time-travel stories. For example, the whole point of the Back to the Future movies is that what time-travellers do can change the course of future history. (If you need, or even want, a refresher on what happens in the movie, one is available here, though be warned that site launches a very annoying MIDI file unless your browser is configured to block that kind of thing.)

In Back to the Future in 1985 the first time around George works for Biff, and the second time around, after Marty has changed the past, Biff works for George. So this is a violation of the one-dimensionality principle. I had always assumed that the movie could be made sense of on a ‘branching time’ model. Indeed in the second movie that’s exactly the kind of model they say they are using.

The idea is that the history we are familiar with is only one branch of the tree of time. This isn’t a wholly unknown picture. I’ve been told that Aristotle believed something similar, and (if you believe everything you read on the web) a few quantum mechanics specialists also hold a similar view. (Personally I think it’s about as plausible as the world-rests-on-a-giant-turtle theory, but the history of philosophers making speculations about physics is not great, so I’ll be a little restrained here.) On this picture the other branches exist, and the only thing that’s special about our branch is that we’re in it. Before a branch point it isn’t determined which branch we will end up on. The full story of the world includes a whole array of things totally unlike anything we know – our history is the story of a particular climb up the tree of time, a climb that could have turned out very very differently to how it actually did.

It should be easy to fit Back to the Future style time travel into this picture. When Marty goes back into 1955 it isn’t pre-determined whether he will stay in the branch from whence he came. And he changes his world enough that he more or less has to move into another branch – ultimately a branch in which his parents are much more successful than they actually are. (Or were. Or something. Ordinary tense words don’t handle this kind of situation very well, as Douglas Adams pointed out somewhere.)

So far so good. Now obviously one part of the movie isn’t compatible with this picture. If Marty is safely and soundly in his new branch, there’s no reason to think he will ‘fade away’ if in that branch his parents don’t meet and marry and conceive etc. He’s there and that’s all there is to it. So a major plot line of the movie becomes a little incomprehensible. But apart from that, I thought it was going to be possible to make sense of it all.

What surprised me on re-watching the movie [2] was that even granting them a branching time universe, and ignoring the lack of reason for Marty to ‘fade away’, the story in the movie still didn’t make sense. Here’s why. In the new branch that Marty moves onto, his parents meet, he is conceived, born and grows up in a successful family, rather unlike the family he remembers growing up in. Marty also travels forward in time in that branch from 1955 to 1985. The Marty that got to new 1985 by time travel is around at the end of the movie – we see his surprise at how different new 1985 is. But the Marty that was born, raised etc is not. On the branching time model, there should be two Martys around now, but the movie only gives us one.

Maybe the movie could make sense on an even stranger metaphysics than regular branching time. What we need is a metaphysics with not only branching time, but also some cross-branch relations that determine who (in one branch) is the same person as whom (in a different branch). And we need those relations to have enough causal force that when a person is in a branch they shouldn’t be in, or are too often in a branch they shouldn’t be that often in, the relations somehow make the world fix things. But even this doesn’t explain why new 1985 Marty should not remember growing up in a successful household. It’s really all a mess, even granting a really wild metaphysical picture. What amazes me is how it seems to work under its own logic while one is watching it. Some enterprising grad student should work out just what that logic is – they could probably justify anything whatsoever using it.

[1]There are several interesting aesthetics questions related to this distinction. For instance, is it a vice in a time-travel story that it does not make sense on a one-dimensional view of time? I used to think the answer was yes, then I decided that was much too snobby. But after my recent bout of time travel movieing, I’m drifting back to my former position. At the very least, it’s a virtue of those stories that do keep to one-dimensional time, just because one-dimensional time-travel stories are so pretty when done well. The plot devices in the last two Harry Potter stories may have been fairly awful, but the time travel story at the end of The Prisoner of Azkaban is rather good for just this reason. That story gets bonus degree of difficulty points for having the characters interact with themselves (admittedly at a distance) in a more-or-less psychologically plausible way.

I think that stories that violate this constraint too frequently rely on our assumption that causality always moves forward in movie (or book) time. I’d be surprised if someone could tell a decent time-travel story in a movie where the order of scenes didn’t match up with what happened in real time or in any character’s personal time. (Think Pulp Fiction meets Back to the Future.) I imagine that the result would be incomprehensible. I’ve seen some people argue that the final scenes of Tim Burton’s Planet of the Apes should be understood this way, but since those scenes are incomprehensible, that doesn’t really hurt my point. On the other hand, I imagine that with some ingenuity one could chop up a good ‘one-dimensional’ movie like 12 Monkeys into all kinds of rearranged scenes and it still be tolerably coherent.

[2]Well, not the only thing. As has been noted here previously, the 80s were a really strange time. The ‘fashions’ are … well the less said the better. But the thing I’d totally blacked out was that in the movie they try and make Marty look cool by having him play in a Huey Lewis cover band. It’s hard to comprehend what they were thinking. I was rather shocked to hear a Huey Lewis song on a ‘classical rock’ station in Seattle, but the idea that at one time associating with his music was a way to impress pretty 17 year olds is just wild.

On the other hand, I shouldn’t play up the fact that I remember much of this time at all. Many of the students in my course won’t have been born when Back to the Future was released. Hopefully that means they won’t ask too many hard questions about why the plot doesn’t seem incomprehensible on first viewing.



Dick Thompson 08.29.03 at 2:27 am

And how would you describe Heinlein’s classic story “All You Zombies”? Of course it’s not a movie, but it ought to be.


John 08.29.03 at 2:48 am

Heinlein’s story is a one-dimensional consistent time travel story. I’d note that the old (well, not so old, mid-90s) animated TV series “Gargoyles” had several episodes with good one-dimensional time travel premises. That show was pretty damned smart for a Disney cartoon show…

And, yes, Back to the Future makes no sense at all.


Ssuma 08.29.03 at 3:16 am

Not to be anal or anything, but Marty does not play in a Huey Lewis cover band. Huey is the judge who tells Marty that his band is “just too loud” to play the school dance. The only song Marty actually plays in the movie is “Johnny B. Goode.” The bandleader, who is supposedly Chuck Berry’s cousin, calls him (Chuck) up to tell him about this great new song. This brings up the question of where exactly the song comes from, if you like those sort of silly philosophical time-travel questions. Which I guess you do.

I only saw the first movie, but I have fond memories of catching the most pompous professor in my department coming out of BTF III at the theater. Listening to him try to explain his “interest in the level of current popular culture.” is one of my fondest memories of grad school.


Brian Weatherson 08.29.03 at 3:36 am

I don’t have the movie here, but I could have sworn the song he played for the judges was a Huey Lewis song. Not that I _want_ to be an expert on _Back to the Future_.


John G 08.29.03 at 4:01 am

Ahem. Marty plays for Huey Lewis as a judge. But what he and his band play appears to be a thrashed-out version of the opening chords to “The Power of Love,” though this is far from explicit.

If you’re not teaching “Timecop,” your students are missing out. Also, don’t forget that John Woo’s terrible version of the time-travel story “Payback” is coming out.


John G 08.29.03 at 4:04 am

P.S. My favorite time travel story is Philip Dick’s “Captive Market.” It’s about an old lady who drives a truck from her rural California market into the post-nuclear future, where she sells goods to miserable irradiated humans at a premium, and has the power to alter the future for maximum profit.

Speaking of Terror Futures…


Kynn Bartlett 08.29.03 at 6:54 am

Don’t forget “Bill and Ted’s Excellent Adventure.”

I’m serious!



SheriHarris 08.29.03 at 7:07 am

Consider the movie “Somewhere in Time”, based on Richard Matheson’s excellent book Bid Time Return.
The novel, written in journal form, is even more effective. Another well-written novel is Doomsday Book by Connie Willis.


Keith M Ellis 08.29.03 at 7:54 am

Note to self: remove Brown from list of schools to consider for children.


Keith M Ellis 08.29.03 at 9:47 am

My previous comment was unnecessarily harsh. I just find it disapointing that even good schools (and good departments) find the need to attract undergraduates with “sexy” classes like this class on time travel. Time travel, as a philosophical problem, is dependent upon some fundamental concepts that are themselves extremely thorny and poorly understood. I can imagine some aspects of the problem sliced very thin and productively examined using analytic tools, but that’s not the purview of a freshman seminar and would bore them in any event. What attracts those students to a time travel class is the opportunity for a lot of hand-waving (and mind-blowing) that exemplify the very worst of philosophy and validate its popular caricature.

Additionally, your apparent ignorance of and disdain for a physical treatment of time travel is itself damning given that QM and Relativity have the only rigorous things to say on this subject (time, causality) in the history of philosophy and science. Yeah, the many-worlds interpretation is hand-waving, but then so is the Copenhagen interpretation and in any case, as I said, QM is relevant. I don’t completely agree with Nietzsche’s comment on the irrelevancy of philosophy after the advent of science, but this example supports his case. What, really, is going to be accomplished in this class?

If you’re going to present thorny, unsolved problems to freshmen, why not just offer a seminar on time itself? That’s a fundamental problem of real interest in both philosophy and science, and a presentation of how this problem has been approached in the history of philosophy (and science) would serve as a nice introduction to philosophic discourse. And it would still be interesting and attractive to ADS-afflicted students.


jbgust 08.29.03 at 12:24 pm

let me second the motion on bill and ted’s excellent adventure. near the end of the film, when the heroes realize that they need something, they decide that in the future they will travel back in time and leave it for themselves to find. they then go to the spot they plan to drop the item and discover that it is there.

i always wanted to have that power.


Kieran Healy 08.29.03 at 12:30 pm

Hey Keith -

Of course it’s possible to take a philosophical approach to time-travel and treat it in a rigorous and interesting way (my wife teaches a course with a component like this in it). Physical theories aren’t immediately relevant because you’re teaching the students about logical possibility and consistency. These are central topics in metaphysics; and winkling out the issues teaches students how to think straight.

QM and Relativity have the only rigorous things to say on this subject (time, causality) in the history of philosophy

Here ya go.


Brian Weatherson 08.29.03 at 2:24 pm

Thanks Kieran!

Actually much of the course will be about general issues about the role of time in classic philosophical problems – like whether the past and future are real, what (if anything) is wrong with the argument for fatalism, whether free will and determinism are compatible, how many concepts of possibility we have in our folk conceptual scheme and whether we can say something substantive about personal identity across time. There’s plenty of other philosophical issues we could get into – the nature of causation, the aesthetic issues raised by incoherent stories, etc. If anything the problem is that the course will end up being too unfocussed, too much like a week-of-everything-intro course. In any case, I certainly don’t feel bad about making some effort to make these topics interesting to as many students as possible.

I am fairly dismissive of the many-worlds interpretation (largely because I’ve never seen a good account of how we generate the _probabilities_ distinctive of QM in the many-worlds account), but really I don’t think it’s particularly relevant to time travel. What might be relevant to time travel are Godel’s models for general relativity that include closed time-like curves. Of course, it doesn’t look like our world is particularly like any of those models, and time travel in such a world would be, at best, prohibitively expensive, but it is at least relevant in principle. Still, it’s hard to see how the existence of such a solution to GR is relevant to the issues about fatalism, free will, personal identity, etc. that are the main _philosophical_ focus of the course.


Brian Weatherson 08.29.03 at 2:30 pm

PS: I agree _Bill and Ted_ is a fairly interesting example. I couldn’t really make sense of the thing about Ted’s watch. I think what happened was there was only one time per day they could travel into. (Or perhaps it was just incoherent.) But the scenes at the police station are clever, and a nice illustration of what you can and can’t do using time travel in one-dimensional time.


tim 08.29.03 at 2:41 pm

Keith – don’t be so quick. The “note to self” was not too harsh. I was at Browm (for a grad degree), and would never, ever recommend it for an undergraduate education. It is a University that has consciously made the decision to trade academic rigor for a summer camp atmosphere, on the belief that graduates will remember the latter more fondly and demonstrate that through their donations. To be fair, it is once again “under new management,” but it will take a while for the effects to wear off and the boutique courses (“Hard Choices” et al.) to disappear even if the new leadership charts a different course.


SKapusniak 08.29.03 at 2:55 pm

‘Disappearing Marty’ is a big problem, however I don’t think ‘Missing Marty’ particularly is.

Where is ‘Missing Marty’ then? Well he’s back in 1955 having travelled back there in a time machine built by a ‘Missing Doc’, or in some other 1985 having come back from 1955.

So why in 1955 do we not see multiple Marty’s swarming around, having arrived from many many different 1985′s?

Well the point about a branching model is firstly it didn’t start branching from 1955 it started branching from the beginning of time, therefore there are many many 1955′s out there to choose from when travelling into the past. Seoncd by it’s nature ‘time-traveller arrives’ is presumably an event which causes branching.

Throw in a rule that Doc’s time machine can only take you to a future that is consistent with the past/present you’re currently in, which seems pretty plausible on the evidence shown to us, and we have no reason to believe that it’s any more likely that ‘Missing Marty’ would end up back in the same 1985 he started out in than it is for ‘non-Missing Marty’…

…and I didn’t even have to bring in the fun stuff like multiple convergent pasts. Curses!


Andy Ruff 08.29.03 at 4:26 pm

May I suggest watching Donnie Darko. While following a many-worlds theory, it has a couple interesting insights into perhaps the choices one would make given the knowledge of time travel.


mottyl 08.29.03 at 5:04 pm

Gene Wolfe in ‘Free Live Free’ tries to deal with the ‘interacting with yourself’ problem; I’ll leave it to you to decide whether it’s done well or not.


arthur 08.29.03 at 6:12 pm

Sounds like a cool class. When I have room in my schedule next year, I’ll travel back in time to audit it.

Re: Back to the Future. The best line is a throwaway about Miami winning the World Series in 1997, which did in fact happen, even though Miami didn’t even have a baseball team when the movie came out.


Matt McIrvin 08.29.03 at 6:17 pm

Way back in 1994 I wrote a little taxonomy of time-travel stories, including the ones I thought were illogical (the Back to the Future series was definitely in that category, though I understood that as light sci-fi comedy it was perhaps not properly held to high logical standards). Other people followed up with some variants I had missed.


Timothy Burke 08.29.03 at 6:22 pm

Time travel stories I take to be, and yes, occasionally teach about (strike Swarthmore off the list too, I guess) as a kind of thought-experiment about causality, and they’re kissing cousins in that respect to more formal counterfactual writing in history and other social sciences, and even to social scientific exercises in model-making that isolate a single variable and “test” its importance by subtracting and adding through manipulations of data sets.

Time travel stories are in that respect no different than the kind of (absolutely crucial) philosophical and theoretical discussions that begin with “Suppose we had no senses whatsoever: what would we know about the world?” or “Suppose a man was stranded on a desert island with no human companionship: what would he become?” or even “What were people like before they lived in settled communities?” This is the kind of thinking without which Western philosophy, historiography, literature and science could not have existed. You could imagine a time travel story where the time traveller travels back and eliminates the idea of the time travel story and that might change everything!

To deal with the issue raised about Back to Future, though, that type of time travel story commonly uses a “perspectival exception”, meaning that the traveller himself is exempted from being enmeshed in the changes his actions introduce into the world by virtue of his participation in those actions.

There’s a variation on that formula where there is a sort of amnesiac wave that catches up with the traveller once he returns to a changed present, eradicating his awareness of the transformation, but this is commonly used only when the time travel is “restorative”, meaning it is undertaken to move events back to a status quo state after some other traveller has moved them onto another causal branch.

What I would say is that you’ve identified something that is perishingly rare in the entire genre of time travel stories, which is a traveller who returns to a changed present and is not changed himself where the consequences of the traveller’s lack of being “native” to the new branch of time are then explored.

E.g., it doesn’t strike me as intrinsically weird that Marty should return to a new branch of his present and still be the “old” Marty. What strikes me as weird is that the “old” Marty should be able to live comfortably in the “new” branch–every detail of his life within his family should be completely unknown to him, and in very short order, he should be exposed as a kind of imposter or changeling, or an amnesiac.

Very few time travel stories explore those kinds of issues. Which in the end has to do with the underlying hypothesis that a story like Back to the Future is making about causality: it proposes that its protagonist’s character at the *outset of the story* developed individually, an aberrational exemption from his family and surroundings, and therefore, a change in his surroundings does not change him, it only increases his comfort level within the world. Marty is already outside the causal loop of his world before he ever travels in time.


Matt McIrvin 08.29.03 at 6:22 pm

And if you’re wondering about the Babylon 5 reference in the above, it was to an episode called “Babylon Squared,” and it was revealed a few seasons later that what I thought was happening was not in fact what was happening. It was a nice fake-out. The time-travel story still wasn’t completely logical, but it was clever in some ways.


raj 08.29.03 at 6:31 pm

I’m not familiar with Heinlein’s story “All You Zombies,” but I can attest that his story “By His Bootstraps” that he published under the pseudonym Anson MacDonald is excellent.


tom strong 08.29.03 at 7:07 pm

Marty goes back in time, kills his ‘mother’, his ‘father’, and Biff for good measure. All Marty’s in all branches could then have never be born. Please read your Asimov.


Anon 08.29.03 at 8:11 pm

My favorite time travel story is one from (outer limits or twilight zone, don’t remember which) where a woman goes back in time to kill hitler. She becomes the nanny for the hitler family, kills the baby hitler. Then, she goes out and gets a kid from a poor woman on the street and replaces the original baby.

Well, this new baby is hitler and the other baby woulda just been a normal guy.

I think that is the most coherent type of time travel, where there aren’t two different timelines.


daniel 08.29.03 at 8:33 pm

My favorite time travel story: seeing Huey Lewis in the airport in Salt Lake City.


Duncan 08.30.03 at 1:30 am

John Wyndham wrote a number of good time travel stories: “Opposite Number” from the “Seeds of Time” collection explores the branch theory (using the comparison with quantum mechanics) and “Chronoclasm” has fun with the whole idea of inventions that are never actually invented – because someone from the future shows the “inventor” how it’s done, using information that originally came from the inventor.

I didn’t know anyone had actually made a case for Terminator 2 making sense – I always felt the difference in coherence between that and the first film showed how success has allowed James Cameron to get away with increasingly nonsensical plots.


David Mackinder 08.30.03 at 1:46 am

another interesting movie, recently shown on the BBC, is Gregory Hoblit’s _Frequency_, starring Denis Quaid and Jim Caviezel — though it requires that you don’t really probe too deeply the games it plays with time and alternative outcomes.


Geoff 08.30.03 at 2:04 am

Bonus points to flicks like 12 Monkeys for taking a decent intellectual stab at time travel, but a close viewing of even that film reveals logical inconsistencies that cripple its approach (of course, throw in the suspension of disbelief you need for just about any work of fiction, and it’s a great watch – better than the vast majority of stuff out there). I have yet to see a fully fleshed-out theory of time travel at all, let alone one not reliant on physics sleight-of-hand. Frankly, having studied some physics and QM in university, it seems to me that modern cutting-edge physicists often have little real comprehension of the processes underlying their research, instead substituting analogies and metaphor as alternatives more palatable to the layman’s understanding. It raised the disturbing question of whether or not the physicists really understood what was going on, or whether they simply attempted to grab the first working system that floated by.


karen 08.30.03 at 2:19 am

Brian, One of the things that frustrates me about the otherwise interesting subject of possible worlds/time travel is the way the visions fail to capture the enormous numbers involved, in two ways.

(1) In Investigations by Stuart A. Kauffman (excellent and enjoyable, btw) Kauffman describes and calculates the number of “adjacent possibles” – take the state of the world and add, subtract or delete one atom. Although we know the numbers will be large, they are really different orders of magnitude larger than we expect. Now, this doesn’t pose a problem for philosophy or literature directly, but a model that ignores the most mindboggling part of alternate worlds is esthetically unsatisfying. Worth a mention in your class.

(2) The stories fail to convey how amazingly grainy and sensitive it would be to travel to any one point by following the branching tree. I can accept that every slight variation exists, but changing one person slightly would make enormous and unpredictable changes in subsequent events. Marty would need to balance out all the changes he didn’t want, probably in unrelated ways, to get to the good 1985 counterfactual. Terry Pratchett, of all people, actually addresses this, by saying that, in his universe, the logic of stories is real and exerts a pressure to keep subsequent events as close to the original as possible, even when a change is imposed from outside.

(3) I agree with the suggestion to use Heinlein’s “By His Bootstraps” – it is short, entertaining and clever.

(4) A friend had a lot of business travel, nearly every other day in one month, and kept seeing the same movie on the airplane. The movie was “Groundhog Day” – ’nuff said.


Shai 08.30.03 at 4:04 am

I’ve never been to Brown but i’d like to play devil’s advocate in response to tim’s anecdotal report on his experience there.

Brown is a respectable 17 on the annual US News report, and according to Brian Leiter’s Philosophical Gourmet it’s tied with Yale, Cornell, and The University of Chicago for 16, at 3.4 in its overall ranking of American philosophy departments.

To give an idea of what that means, I am a student of the University of Toronto which has an overall rank of 3.5 (pretty much the same as Brown, but for reasons peculiar to the method of the report probably worse) and next semester I’m taking classes with both Ian Hacking and Thomas Hurka, two very good philosophers. I expect the case for students is the same or better at Brown.

And if the US News Ranking is indicative of anything, Brown in general is certainly not in the same league as Timbuktu college; not that anyone was implying that, but you get the point.


Brian Weatherson 08.30.03 at 4:35 am

There’s lots of good suggestions here. Thanks to everyone, and keep ‘em coming!

I probably should try and defend the honour of my university/summer camp, but instead I want to expand upon my throwaway comment about T2 because it relates to some puzzles about the nature of fictional representation that I’ve been intrigued by.

I agree that the most obvious reaction to T2 is that it’s incoherent, and that it probably should have been read as a warning about James Cameron’s career arc, but it has a really simple coherent interpretation. (Warning: the following will make it painfully obvious that I’m not a literature expert. C’est la blog.)

What seems to make it incoherent is that we’re told that Miles Dyson invents SkyNet, and then we see him die well before SkyNet could be invented. But we never _see_ him invent SkyNet, the main evidence for that is that Arnie says he does. So there’s an easy way to make the story consistent – Arnie simply has very mistaken beliefs about how SkyNet was invented, since obviously it was invented without Dr Dyson around.

This ‘interpretation’ raises a few questions. Why does Arnie have these crazy false beliefs? Does the T-1000 have the same beliefs? If so, why is he so mistaken as well? If not, why does he follow them to Dyson’s workplace? These are all hard questions, but they have answers that are at least _coherent_.

This is the kind of issue about time travel stories that I find most fascinating. You’d think (or at least a philosopher would think) that when we’re given really good evidence that what a character says is false, we’ll infer that the character either has false beliefs or was lying. But in T2 we don’t do either of those things – instead we infer the world is somehow incoherent. Why isn’t the simple explanation, Arnie’s history circuits are all screwed up, the preferred one?

Two reasons spring out. First, in interpreting stories there’s an important phenomena (that I assume has a proper name) that sometimes characters can be framed so that they are taken to be speaking canonically. The character speaks with the voice of the author, either temporarily or permanently. It’s hard to say how or why this happens, but we usually recognise it when we see it. (Having a long monologue is a movie with relatively little talking is one way to do the framing I guess.) And when Arnie is explaining how SkyNet was built, he’s framed in just this way. So saying Arnie got it wrong implies that the author got it wrong. But authors don’t (usually) get it wrong about their own story, because it’s their story, so they get to say what happens.

Still, when an author verges into incoherence we normally pull back this licence. If I start a story with, “Osama, the caretaker’s son, had his character quirks. For example he wanted to murder everyone who wasn’t a Muslim. But all things considered he was basically a decent chap.” I imagine you’ll (quite rightly) object. If Osama wants to murder 5 billion people, he’s not a decent chap, even in a fiction. And you might think that when the author says Miles Dyson invents SkyNet two years after he is blown up in a factory explosion, similarly we would say the author has gone beyond the boundaries of poetic licence.

But in time travel stories we don’t say anything of the sort, and this seems to matter to how we react to T2. I guess that the reason for that is related to the reason we don’t object to the questionable coherence of _Back to the Future_, but that’s not really explaining what’s going on.

(Disclaimer: I haven’t seen T3, so I don’t know whether this undermines the interpretation of T2 I’ve provided here.)


WillieStyle 08.30.03 at 5:11 am

Sounds like a cool class. When I have room in my schedule next year, I’ll travel back in time to audit it.

Why would you need room in your schedule?


Matt McIrvin 08.30.03 at 6:38 am

Geoff: Mathematicians like to accuse physicists of lack of rigor in their theoretical derivations, and they’re right. The reason physicists can usually get away with this is that they’ve got to make contact with real data eventually; the world has the final say. I think the reason physicists haven’t come up with a fully-fleshed-out theory of time travel is that there’s no evidence of time travel to describe with the theory, except for one-way travel into the future that isn’t particularly paradoxical. Most physicists aren’t going to spend much time on a theory of stuff that only happens in science-fiction stories.

Though some have anyway, because of the odd corners of theory that impinge on this. As for Brian’s musings on the relevance of many-worlds to time travel: David Deutsch has long claimed that you can make sense of branching histories in a general-relativistic spacetime with closed timelike lines, like Gödel’s, by applying many-worlds QM to it. I’ve never been quite sure whether I believe him.

Others have been trying for some time to come up with a theoretical basis for a “chronology protection conjecture” under which closed timelike lines wouldn’t ever occur in realistic situations. I remember hearing an interesting talk by Roman Jackiw years ago about a spacetime with closed timelike lines as a result of a pair of cosmic strings: he’d shown that setting this up would take infinite energy, and wondered whether this might be generalizable.


Bruce Baugh 08.30.03 at 10:46 am

I have a theory, based in part on a general sense of Terry Gilliam’s approach to these things, that 12 Monkeys does not present a one-dimensional timeline. It’s at least possible that the man we see in the yellow coat in memories early on is not the man we see in the yellow coat in present time at the end – that the time traveler succeeded in changing the past in ways that neither he nor anyone else ever recognizes.

The anime series X, now up to six discs out of, um, thirteen or so, has some very nice stuff with prophecy and destiny, and the limits on each. It’s supernatural, of course, rather than strictly scientific. (Yes, I do realize this is far too long for a class.)

For the thorough-going mindfuck, there’s always David Lynch’s film Lost Highway. Now of course it’s got a literally incomprehensible moment smack in the middle, but it seems to play out quite smoothly all around that if it’s taken as a given.

Minority Report is worth mentioning if you bring up the possibility that, just as we were told the prisoners all do, the protagonist is dreaming in his confinement of the perfect or at least an acceptable outcome.

Donnie Darko’s a delight, and one of the most thoroughly Phildickian movies I’ve seen.


spacetoast 08.30.03 at 3:58 pm

“whether the past and future are real, what (if anything) is wrong with the argument for fatalism, whether free will and determinism are compatible, how many concepts of possibility we have in our folk conceptual scheme and whether we can say something substantive about personal identity across time.”

It made an awful movie, but Slaughterhouse Five seems like it would be an almost ideal springboard to this bunch you mention. That fairly recent thing with Tom Cruise deals with time by way of causality issues. Err…Peggy Sue Got Married? (shut up)


G C 08.30.03 at 4:31 pm

I’d definitely like to second the recommendation of “Slaughterhouse Five.” I’m actually teaching it this semester in a course that has nothing to do with time travel, but for a time travel course, it’s spot on.

12 Monkeys is also a necessity. In terms of consistency alone, the best time travel movie ever made.

I’d like to comment briefly on the Missing Marty problem. You might not remember, but the “missing Marty” is seen going back in time to 1955 at the very end of Back to the Future Part I. So there should be only one Marty at the end of the movie, because the one who grew up in that timeline is gone.

But there’s still a problem. If the Second Marty can’t possibly do everything that the first Marty did because he starts with different information. He doesn’t know that Biff is a bully who runs/ruins his dad’s life, and he doesn’t know the original story of how his parents yet. (He doesn’t know the birdwatching story, for instance.) Because of this it may not even occur to him to intervene until it is too late.

BTTF therefore gives us a sort of endless loop where a Marty goes back and time and mucks around in 1955, only to return to the future moments before the New Marty heads back in time to change everything around again.

Marties would spring up into existence and then replace themselves forever until finally the universe hit upon a successful, self-causing, self-consistent pattern.


G C 08.30.03 at 4:32 pm

Best time travel webpage on the net, hands down:


Paradox discussion and exploration.


Tom T. 08.31.03 at 4:17 pm

Connie Willis has written a couple of time-travel novels using the notion that time travel is achievable, but a physical property of the universe prevents actions that would foul the timeline. Thus, a traveler by definition cannot get close enough to major events to affect them (in one book, researchers trying to visit the bombing of Coventry keep winding up in turnip fields outside the city at several days’ remove). History, in her conception, is robust, with small perturbations collapsing back into a timeline that is unchanged overall.

Also, another consequence of her theory of time travel is that time travelers cannot bring objects forward from the past. Because of this, after an initial flurry of interest in the technology, the corporate world gave up on time travel as a money-making proposition, so that it is now wholly the province of underfunded university researchers. The crowd on this weblog might particularly enjoy that aspect.

Her first such book is Doomsday Book, about a creepy and tragic visit to England at the outset of the Black Death. Her second is To Say Nothing Of The Dog, an immensely funny Victorian-era comedy of manners, which actually hinges on the fate of a cat. The restraints built into her theory of time travel make for intricate (and generally coherent, I think) philosophical and practical puzzles, especially in the second book.


adamsj 08.31.03 at 11:39 pm

A minor nit to pick:

While the events of “All You Zombies–” are one-dimensional, the backstory is not. The organization for which the narrator works has as its purpose going back in time and preventing or causing historical events.


Brian Weatherson 08.31.03 at 11:56 pm

This bit gets awkward, but it is consistent with 1D time that a time traveller could go back and cause or prevent an historical event. It’s just that it must be the case that the even they cause (prevent) was (not) part of the history they were already in.

If there was an oddly knowledgeable kinda nerdy red-haired Australian who played a crucial role in stopping the Cuban missile crisis turning into WWIII, then I could go back in time and be that very Australian. And since he prevented WWIII, that means I could go back in time and prevent WWIII. Indeed I better, because if I hadn’t there would have been a nuclear war in the 1960s, and we wouldn’t be here. (If you prefer you could go back and lobby for all the treaties that led to WWI. I might try and kill you though if you’re going to, just on general anti-WWI principle. Whether I succeed will depend, inter alia, on whether it really was you who did that 1910s lobbying.)

It’s really hard to think about these cases, because all of our normal ways of thinking about causation involve it propagating forward through time, and the idea of time travel is that there can be backwards causation.

Some people think this means time travel in 1D time is incoherent. (See, for example, the site gc mentions, which has some rather weak arguments to that conclusion.)

I’m much more inclined to think that, as David Lewis said, the correct conclusion is that a world with 1D time travel would be radically different to the kind of world we normally think we inhabit. Since QM already tells us that the world we do inhabit is radically different from the one we think we do, it’s hard to know just how big a strike against time travel that is.


Sean O'Callaghan 09.01.03 at 2:36 am

I would be interested to know if anyone in the class points out that the fact that no-one has ever met someone from the future and for that reason it’s probably impossible to travel back in time.


Brian Weatherson 09.01.03 at 3:05 am

I was going to spend some time on that argument, but I realised I didn’t have a whole lot to say about it. Here’s two things one might say:

(1) We have met time travellers, they’ve just been very well disguised. There are some books that take that line, but it’s obviously not the most plausible thing in the world.

(2) It turns out the only kind of time machine that is feasible requires you to put some part of the machine at either end of the travel. It’s like a telephone – you can talk across long distances but only if there’s a machine of the right kind at the other end. When the first of these machines is built it will be possible to travel back to it, but since there is no such machine in 2003, no time travellers in 2003.

Option (2) looks suspiciously ad hoc as a solution to the ‘no time travellers’ argument, but I’ve met a few people who take it seriously.

As for what it reveals about the students who bring it up – perhaps just that they read the textbook because it has a moderately long discussion of this puzzle, including stories that follow option (1).


Tom T. 09.01.03 at 6:45 pm

There have also been time travel stories that posited that the amount of energy required to travel backward through time rises asymptotically toward infinity the farther back one tries to go, so that a time traveler could never go back farther than, say eighty years from the invention of her machine. In this scenario, we are just too far in the past for any time traveler to have gotten here.

Yes, there are obvious problems with this explanation; I’m just alerting you that it’s been raised.


Chad Brooks 09.02.03 at 9:19 pm

If the course is hot limited to movies check out Poul Anderson’s time travel stories.

Jack Finney’s first time travel book Time and Again is also a must (but not the sequel).

Comments on this entry are closed.