Home > Film > Source Code (2011)

Source Code (2011)

For once we’ve got a reasonably intelligent science fiction film rather than an excuse for poorly realised spaceships to dodge and weave about the screen, firing off superweapons and exploding in balls of fire — something that suggests there’s a previously unrecognised mass of oxygen in outer space capable of supporting combustion. Well, perhaps that’s a slight exaggeration since space opera is moderately rare in the cinema these days. Hollywood prefers the cheaper version of aliens running around, blowing stuff up on Earth. That makes for better explosions and cheaper CGI. Anyway, Source Code (2011) offers us one of the more coherent efforts at a multiverse story even if it’s more than a little amoral.


No, really? A multiverse story? What’s that?


Well, before we get into spoiler territory, let’s deal with a little of the background. This theory suggests we have more than one reality. Conventional physics says time, gravity and all the other constants move in a straight line. So although each individual makes choices, the outcomes to all the decisions are fixed in the one timeframe. We all live with our triumphs and mistakes equally. But others suggest each decision is like a fork in the road. Sometimes we walk left, sometimes right. So in parallel dimensions, we live out our lives with each set of choices. In the most interesting of these theories, there are an infinite number of possible universes because, over time, millions of us make decisions every day and so the number of possible outcomes expands without limit.

Jake Gyllenhaal and Michelle Monaghan try to remember what happened last time


OK, taking this as our base, we assume that, at any one time, there are an infinite number of realities almost exactly the same as ours shading to realities nothing like ours. Dr Rutledge, played with introverted intensity by Jeffrey Wright, has developed a technique for implanting the mind of a man from our reality into the body of a matching man in an alternate reality. If you recall Quantum Leap, we’ve the same convention that Jake Gyllenhaal is transplanted into a different body, but we continue to see Jake Gyllenhaal. This form of the transplant keeps his female fans happy and lasts for exactly eight minutes, at the end of which the host dies (not through the shock of becoming Jake Gyllenhaal, you understand, but in an explosion).


For those who like to play around with the ideas, the death of everyone on the train in this alternate reality prevents there being any contamination of that timeline. Even though the transplanted man may say or do things to disturb the alternate, the effect never leaves the train. Dr Rutledge assumes that if our hero, Colter Stevens played by Jake Gyllenhaal, can identify the bomber in the alternate reality, the same person, driving an identical van with the same number plate can be arrested in our reality and so prevent a second explosion, this time a dirty bomb.

Vera Farmiga and Jeffrey Wright waiting for the next cue


It’s actually better not to think too much about this because the chance of people in alternate realities having the same name or the same licence plate on their vehicles seems remote. I suppose this may occur because the source code recreates a “captured” version of the alternate reality every time the program is run, i.e. it starts with the same parameters every time. If this is the case, the good doctor is creating millions of people in a new reality just so a small number can be blown up on a train and millions can be maimed or killed in Chicago when the dirty bomb explodes. This act of creation and death is justified because it’s expedient to save our people. It would be less immoral if the alternate realities already exist. Now all we’re doing is exploiting what’s inevitable for them, so that we can avoid the same fate.


No matter how it works, in each of the eight minute insertions, Colter Stevens learns about the people in his section of the train. He does this by being prepared to beat them up and, if necessary, shoot to kill. We’re not supposed to care because the people we see only have eight minutes to live. What happens to them is irrelevant in the larger scale of things. In the midst of his investigating, Colter Stevens finds himself attracted to Christina Warren, played by Michelle Monaghan. He decides he should try to save her. This is interesting because, should he succeed, one person surviving the train explosion will produce a major divergence of the realities. That need not concern our timeline, of course. It just means there will never be any chance of going there again as this person now interacts with thousands of people during her lifetime, thereby moving that reality ever further away from ours.


With Colter Stevens dying every eight minutes, he develops psychological problems. Encouraging him to keep going is the pivotal Colleen Goodwin played with quite remarkable sensitivity by Vera Farmiga. Without someone strong in this role, the film would collapse. She’s pitch perfect throughout and gives the film unexpected weight.


This is the stand-out science fiction film so far this year. Jake Gyllenhaal strives valiantly in a slightly thankless role while everyone else, led by Vera Farmiga, rallies round and produces an excellent ensemble piece. It’s a clever script by Ben Ripley allowing the scenario on the train to continuously evolve and expand. For once, Ripley has produced something better than films about sex-crazed aliens, with the whole thing beautifully directed by Duncan Jones, who seems to be making a name for himself rather fast. All in all, Source Code is excellent viewing for anyone who likes science fiction which follows through to the implications of our actions no matter how immoral.


Stop reading here if you don’t want a discussion of what actually happens.


We get this far by suspending disbelief and accept the arrest of the bomber in our timeline. Not being sure how the machine works, we may have to thank Dr Rutledge for destroying Chicago in perhaps more than one hundred other realities depending on how many times Colter Stevens iterates through his eight minute loops. But we are safe. Our Earth’s authorities are delighted with the outcome and can’t wait to use the machine again. Before they embark on new threats, I sincerely hope the Government intends to use the machine to save as many alternate versions of Chicago as possible. This would be the moral step, maximising the benefit of this invention for all realities. We would want other realities to save us if they could, so every Dr Rutledge should be arguing for his Colter Stevens to help others before he helps himself. Sadly, we see Dr Rutledge rubbing his hands and only speculating on what his next triumph will be, confirming the general lack of morality in this project. This is selfishness personified, a sauve qui peut approach to life.


Perhaps anticipating how he will be used and taking everything he has learned about the train, Colter Stevens now knows enough to prevent the train from blowing up. He therefore persuades Colleen Goodwin to send him in one last time to save at least one Chicago. At the end of this eight minutes insertion, she’s to turn off his life support and let him die. This she does. The result is presumably an arrest with her sent off to die in the Marine Corps Brig in Quantico.


After the freeze frame, we are in the alternate reality where Colter Stevens saves Chicago, gets the girl, and sets off for a new life with her. What makes the ending initially appear so pleasing is the text message he sends to Colleen Goodwin in this new reality. For, yes, there’s an identical project in this reality with a version of himself waiting to be deployed to solve a major crime and avert catastrophe. This message primes Colleen Goodwin to encourage Colter Stevens. Not only can he “save the day” no matter where he’s sent, but he can also escape and find a new life for himself in an alternate reality. So each reality may be said to offer Colter Stevens hope, no matter how desperate things may seem. No-one can ask for more than that in any reality. Let’s not go into whether our hero could sustain a convincing impersonation of a man in that reality, once it’s confirmed he can stay with the new identity. There’s also an unresolved paradox because, if Dr Rutledge’s technology depends on the target man dying, he no longer dies, i.e. the transfer should not work.


Now let’s come to the really big question. Colter Stevens knows he displaces the mind of the man in the target body. Let’s say he believes the mind of the teacher is transferred into his body. When he persuades Colleen Goodwin to switch off the life support, he intends Goodwin to kill the teacher so that the replacement is permanent. In my book, that makes Colter Stevens and Colleen Goodwin murderers. However, no matter what he believes about where the mind of the teacher goes, the clear intention is to kill that mind so that our “hero” can have a happy ending. There used to be a morality code in Hollywood. It was known as the Hays Code. Although this was predominantly concerned with sexual and, to some extent, political content, there was a general view that motion pictures should not show criminals benefitting from their crimes. Under the Code, this ending could not have been added after the freeze frame. The rule used to be that criminals should be punished. While this is, no doubt, unacceptably black and white for our relativist age, I’m surprised a stone-cold killer should be shown enjoying his stolen life in the final frames. I’m not sure what message this is sending to our impressionable young.


Source Code (2011) has been shortlisted for the Ray Bradbury Award for Outstanding Dramatic Presentation 2011 and for the 2012 Hugo Award for Best Dramatic Presentation — Long.


  1. PRueda
    September 12, 2011 at 10:41 am

    As you have stated, they are using the concept of Alternate Timelines. The concept that for every action we do, a fork in the road is created with the results of what would have happened if we had made the alternate decision. This can get complicated because due to the butterfly effect, even small insignificant things that don’t seem to make a difference in real life will throw our alternate selves off by milliseconds which will turn into seconds and minutes over time and create a huge change in the timeline as it progresses away from the point of divergence.

    So, here we have a train Bombing. In the initial timeline, the first one before Source Code ever enters the picture, the bomb explodes and everyone dies. In Timeline A, Sean Fentress remains himself and is not affected by Coulter Stephens… this is the timeline where Goodwin, Rutledge, “the time capsule” timeline exists… Timeline A. This is where the Science Fiction aspect pops in… has Rutledge created a program that can thrust someone into an alternate reality? Yes!

    Rutledge’s theory is that the program is manifesting some brain waves into a computer program and it is placing Coulter Stephens into the program where he can use the details in the brain waves to figure whom the bomber was. Except… he isn’t doing that at all. If this was the case, then Coulter wouldn’t be able to control anything Sean does because Sean already lived those memories and they cannot be altered. What his program is actually doing is creating a new series of Alternate realities that diverge at the point they send Coulter into the Source Code. The timelines they create contain all the same exact details that occurred in the world of Timeline A up until the moment Coulter enters the timeline, including all decisions, actions, etc by everyone in the world. The moment he enters, though, the timeline diverges away from Timeline A and creates Timeline B…or C… or D (from our perspective). So now, Coulter has created a new series of events in a different universe that is identical to his original universe, except for his entrance into it. Since in most of them the train explodes and all those he affects die just like they did in timeline A, we can assume that his presence there has little effect on the timeline as he will die and so will everyone who interacted with him… with the exception of the two timelines where he gets run over and the one where he confronts Frost. In those two, Christina lives in one and in the other one the timing of the bombing was off so it probably didn’t coordinate with the tanker train and didn’t kill as many people. In effect, my point is that every time he returns to Timeline A, he has not let go of the timeline he created, it will continue without him in it because he technically dies every time (bombing, getting run over, or getting shot). So, Coulter is actually going “back in time” and creating a new divergence point from his own Timeline, or a new alternate reality. But the program is only allowing him to go through this new reality 8 minutes before it brings him back… he figures this out eventually.

    One can also assume that every time Coulter fails and the train explodes, more and more timelines are created from that divergence because that new timeline will have another Coulter Stephens who will go back in time and do the same thing… infinitely (but again, our perspective as the observers is focused only on timeline A, so we don’t go past that). This makes the most sense because if Coulter really were just going back into the memory of Sean Fentress, he wouldn’t be able to see or know where anything or anyone that Sean didn’t interact with is, what they did, etc. (like, for example when he exits the train after the Arab guy… unlikely Sean could’ve retained memories of the train station because Sean didn’t get off the train)… so the only viable reasoning for what Coulter is able to accomplish in the Source Code is that he is continuously being sent back into Timeline A at 7:40am, and creating a new timeline where the events are exact up until the moment he arrives when they diverge as he interacts with them… (this is why Christina always says the same thing when he wakes up, and the soda can always pops, etc, he doesn’t arrive far enough in time to affect these things so they are always the same).

    In the end, he wants to save the train and he does it… when the 8 minutes run out and Goodwin shuts off life support, Coulter has nowhere to return to, the program is no longer affecting him because when he died he was disconnected from it and so he remains in that timeline he created by saving the train and bringing Derek to justice… he created the timeline when he became Sean Fentress for the last time, so Sean doesn’t exist anymore after that divergence, he will instead continue to exist in the opposite diverged timeline where he doesn’t become Coulter and dies (technically, timeline A). So Sean is always doomed (kinda sucks for him). Because of this, Coulter is technically Sean after this point as far as the universe sees him, meaning there is no paradox with two Coulter’s existing in one universe, but his consciousness remains that of Coulter since that is who he is in Timeline A which is where he came from. So, he has been given “another chance” to live through Sean’s life… a life that would’ve been doomed to end anyway since Sean will die unless Coulter intervenes (reasoning for why Goodwin and Coulter don’t murder Sean Fentress when they do this, because Sean will die anyway, unless Coulter intervenes).

    The only thing now is that… Coulter has no memory of Sean’s life after he entered his body and created this new reality (which is like Timeline M or O since he has gone back so often), he won’t know his address, friends, families, nothing… which is going to be challenging, but I guess better than where he was before.

    • September 12, 2011 at 5:16 pm

      This seems a reasonable explanation of what’s happening although, as I commented, this means a new timeline where the city is destroyed is created every time the program iterates. The price of saving Chicago in what you term Timeline A is that Cutler goes through the cycle enough times to learn how to stop the explosion. So morally, is the Rutledge solution appropriate, i.e. destroying so many alternates so that the one Chicago can survive? Personally, I think the deaths of millions in all the alternate timelines is too great a price to pay but, if I were Rutledge, I might be more cold-blooded.

  2. PRueda
    September 13, 2011 at 8:21 am

    Whether it is immoral for Rutledge to willingly create a new universe and place its inhabtants in danger to benefit his own would depend on whether Rutledge even understands he has created a new alternate universe at all… is it immoral to destroy something you are not even aware exists? It also brings into question whether the well being of another universe is even a concern in ours, they don’t exist in our reality, so they are not “real” and as such, are the same as asking whether killing animated beings in a video game is morally wrong, too?

    • September 13, 2011 at 10:47 am

      Now that’s a fascinating idea and it captures the central paradox of the alternate timeline theory. Let’s assume Rutledge understands how his “machine” works, i.e. that every time he sends Coulter back to the start of the sequence, he knows that a new “world” is being created. He’s attempting to save his own “world”, or Timeline A if you prefer, by these repeated sendings, so he must believe his “machine” is recreating a real world and not a purely electronic version as in your video game analogy. His “world” cannot benefit from a purely theoretical exploration. Each iteration must be an exploration which has physical consequences in his “world” otherwise there’s no point to the experiment.

      As long Coulter’s explorations do not change the explosion on the train, all the minor changes end in Timeline A. But the moment Coulter interacts with people off the train, i.e. by making a telephone call or getting off the train, the timeline inevitably changes. Now here’s the problem. If each iteration creates a physical reality there are two possibilities. At the end of the sequence, everyone in the recreated “world” dies when the “machine” completes its cycle. Or each new “world” continues as an alternate timeline and only Chicago is repeatedly destroyed. I cannot see any other way for the “machine” to work. Since there must be an effect experienced in Rutledge’s world, I prefer the second explanation. If the created world is terminated at the end of eight minutes, how would be benefit of the changes spill over to Rutledge’s world? Yet this leaves a paradox. For Rutledge personally to feel the benefit, the “machine” must be moving his whole world back to that one point in time. That’s like resetting the clock. So presumably, the alternate world continues after the “machine” is activated and his world goes back in time. Now, if Coulter is successful, he will pass on the benefit to Rutledge’s world. Think of the machine as being like a lifeboat. When you can see the iceberg coming, Rutledge jumps in the boat and tries to row away from danger. The iceberg still hits the Titanic in the first version of reality, but everyone in his boat has the chance to be saved. For all the first iterations, it does not work. Everyone still dies. When Coulter succeeds, Rutledge is preening himself and looking for the next disaster to avert. He’s proved he can change the world.

      In moral terms, it’s the difference between killing everyone in the “world” every time the “machine” iterates and only killing Chicago multiple times. If all Rutledge was doing was playing a first person shooter game with Coulter as the shooter, he could switch off the machine and his world would be completely unchanged every time. Electronic games do not affect our world. Therefore, he must know he’s killing millions, if not billions, of people.

  3. Steve Watchorn
    April 16, 2012 at 10:53 pm

    The fascinating thing about the film is that it does not really explicitly state what the source code is doing, but if you understand it in certain ways, it is pretty neatly self-consistent.

    In the lack of a clear explanation, I would disagree with the analysis above as to what the source code is doing. I agree with the idea of alternate branching realities as a consequence of a quantum mechanical view of time and action. However, I do not believe source code is “creating” any realities. I think of the time tracks as this film presents them as a nest of wound and jumbled wires, like in a junction box, except there are infinite numbers of wires looping in and around, and there is no concern about “filling up space,” so that an infinite number of wires can be jammed together.

    What you called Timeline A above is just one of those wires. Given all the infinite possibilities of choices that could have been made throughout the history of the universe, it is pretty much guaranteed that, for any point on Timeline A, there will be an uncountable number of timelines running “next” to Timeline A, but in those timelines, the events on the train are happening. Seems unlikely? Remember that infinity is a lot. Even if there were a billion billion billion billion billion other timelines (and infinity is well more than that), and only 0.0001% of them met the criteria of running adjacent to Timeline A at a given moment when Stevens is to be sent, that would still be over a billion billion billion possibilities. One of the criteria is that the world developed more slowly by a few hours (however long between when the train in Timeline A explodes and when Stevens is to be sent). So here is what I think source code is doing.

    The technology exists to project a consciousness from one quantum reality track to another, through the “quantum foam” that exists at the subatomic level (as long as the recipient and the person whose consciousness is to be projected are “compatible”). What the source code does is, using Sean Fentress’ perceptions, it can search the parameters of adjacent quantum timelines until it finds one which matches the criteria having the train running at that moment, and having the bomb on it, and meeting any other important criteria that can be gleaned from Sean’s consciousness. It then projects Stevens into that reality from the point Sean last remembers, eight minutes before his death.

    In that way, no one is creating a “new reality.” They are putting his consciousness into a reality that already exists, as full and complete as our own. He can alter that any way he wants. Now, why do they insist then that he can only stay eight minutes? There are two possibilities: Rutledge does not understand the implications of what he has developed (unlikely), or he uses the veil of “you wouldn’t understand” to shield Stevens (and Goodwin, to a degree) from asking uncomfortable questions. He does not want Stevens to know that they are choosing to bring him back after eight minutes, even if he does not die in the explosion (or some other way); and that, if they didn’t, he could do exactly what he does at the end when they finally do let go of his consciousness: just get off the train at the next stop and walk into his new life as “Sean Fentress.” Rutledge needs Stevens to believe there is no other reasonable alternative.

    Just look at how defensive and upset Rutledge gets any time someone questions him. He knows what he is doing, and what some of his underlings (and Stevens) would think of him if they knew as much as he did.

    So, each time Stevens gets sent, it is into a different, already running and fully formed reality timeline which happens to have the critical moments passing adjacent to our timeline (in a quantum sense) at the moment Stevens’ consciousness is being sent. And when he is disconnected at the end, and they cannot bring his consciousness back, he goes off into a new life. The beautiful thing about the ending is that there is also a source code project in that reality (because it is so close to our own), and Stevens is able to get a message to that reality’s Goodwin to tell that reality’s Stevens that “it will be okay” — a repeated line and perhaps the philosophical crux of the film. I could easily imagine that Stevens then going into a different reality and doing the same thing, and so on, and so on, spreading out like a ripple into the water, spreading the message to countless realities in an endless, lovely string. This idea is captured visually by the perfect image of “The Bean” in Chicago. It is a gorgeous ending.

    The only niggling point here is “what happens to Sean Fentress?” Clearly, Stevens does not carry any of Sean’s knowledge (and who knows what will happen if he tries to teach a history class). He appears as Sean, clearly, but where does Sean go? Goodwin at one point cryptically says “he is, for all intents and purposes, you,” notably saying “he is you” rather than “you are him.” Perhaps Sean’s consciousness does not even realize anything has happened. But I am not sure. In all this, alternate-reality Sean (actually, all of them) get the short end of the stick.

    • April 17, 2012 at 12:09 am

      First, many thanks for taking the time and trouble to make such a significant contribution. I particularly like the idea of the code enabling Stevens to search adjacent realities until it finds one matching Timeline A. However, the code appears arbitrarily to interrupt Stevens after eight minutes no matter what he’s doing. At face value, this would seem to militate against your suggestion that Rutledge must somehow bluff Stevens into returning. The returns seem involuntary and, where it not for Colleen Goodwin’s intervention at the end, he would presumably return. However, this impression may be to do with the point of view. If we’re to see this only from the point of view of Timeline A, then we may only have a window of eight minutes into each new reality. It may well be that a copy of Stevens is projected into each matching adjacent reality and we can observe what happens to “him” during that time. Then, for better or worse, the connection is broken and the copy gets to die or walk away in the other reality. This does mean, of course, that in each transfer of Stevens, Sean Fentress is displaced, perhaps permanently. The copy might terminate when the connection with the code is severed after eight minutes and allow the Sean personality to resume. Or, when Goodwin switches of the machine, it may prevent the deletion. The morality is this is probably acceptable if it’s only an eight minute displacement and Sean ends up no worse off, i.e. not arrested and prosecuted following his removal from an unexploded train. But if Rutledge is knowing killing the personality of each Sean every time he initiates a transfer, this is a form of murder.

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: