Movie Review: Source Code

My mom refused to see Source Code. “If they got the title wrong,” she said, “how can the rest of the movie be any good?” As for me, I expected to say, “It wasn’t Inception, but it was pretty good.”

So, it turns out it was pretty good, but definitely not great. On the positive side, it felt nothing like Inception and I wasn’t tempted to compare the two. On the negative side, the technobabble was worse than I expected. TLDR: it’s a premise-driven movie with a story that hasn’t been done before, but as a movie it’s just okay. Better than Adjustment Bureau, worse than District 9 or Inception.

Source Code is the new (?) movie by Duncan James, the guy who made the indie hit Moon two years ago (which I haven’t seen yet). The premise is pretty straightforward for a sci-fi fan: a new project allows a person to relive the last eight minutes of someone else’s life. They put this to use trying to stop a terrorist, who blew up a train on the morning of the movie, by sending someone back to the explosion to identify the bomber.

The main character, US Air Force captain Colter Stevens, doesn’t want to cooperate at first—from his perspective, he’s trapped in a little capsule, unable to contact the outside world, and with holes in his memory from the whole process. (They go through a memory exercise to help bring him back from the experience, which is good.) Colter realizes he can change reality “within the source code”…but of course, as the scientists patiently repeat over and over, it gets reset every time they send him back. (It was fun watching this after playing Portal 2…his capsule felt very Aperture-ish to me.)

A moment for the technobabble. It is laughably bad, with references to “parabolic calculus” and of course the titutar “source code”. And I went to see the movie with five others, four of whom were engineers, so when I say “laughable” I mean “we were actually laughing out loud”. The fifth guy said if you’re not an engineering type, it’s not that bad, but it was really pathetic. Probably the worst technobabble I’ve heard, ever.

Which is too bad, because it’s actually a (fairly?) reasonable premise. Clearly, if Colter can make different decisions when he’s “in the past” then it’s not just a replay of a dead man’s last eight minutes. But the idea is that with quantum physics, they’re making a different universe, based on the imprint of one person’s “last eight minutes”, and putting Colter in that universe. Impossible, but fairly reasonable for sci-fi.

Of course, that universe still ends when the eight minutes run out.

The style of the movie is pretty low-budget (despite being ~$30M). Only three or four locations, no complex gadgets or cars. A couple of nice special effects (there was one shot of slow-motion fire inside the train that was very cool). But there was also something about it, about the script and the timing of shots, that I couldn’t put my finger on. Almost familiar…

I realized, eventually, that what it reminded me of was Windup, a piece I directed as part of a Theatre Rice show. Windup was based on the Philip K. Dick short story “The Electric Ant”. If you click through and watch it, you’ll see an interesting but far from perfect sci-fi story that’s nevertheless much better than my first draft. (Many thanks to my troupe for working through the script to get to this point!)

The problem with the original version of Windup, and sadly to some extent with the final version, is that the story is essentially based on the premise: the main character is a robot. There are then a few plot points that have to be reached (partly because they’re necessary for the premise to grow, partly because they’re in the original), and then you get to the finale. There isn’t much in Windup that’s character-driven, except perhaps local bits of dialogue between the plot pieces.

(This is the most negative I’ve been about the piece ever, by the way, but I hope I’m not getting my troupe down. You all did amazing jobs and gave the piece its life, and I’m still really happy that we accomplished what I had dreamed about doing for months.)

Anyway, I feel like Source Code has the same basic problem. It has a very interesting premise, which is revealed in pieces throughout the movie (in a way I’d describe as “not predictable, but not surprising either”). It has plot points it has to hit. And so the plot ends up driving the movie. This isn’t universal: the characters that aren’t part of the plot have pretty character-driven dialogue (such as the love interest, and Russell Peters’ comedian cameo character). But it doesn’t help enough, because they don’t drive the plot at all except indirectly. The plot itself is too much a train on tracks to change directions in response to its characters.

The timing was off, too. At a macro level, the movie spent a long time on things that ultimately weren’t that important, and other times sped through scenes I actually wanted to watch. This is the harsh time limit on the movie: you can’t show everything, because then you get a ridiculously long movie. That’s not just a practical consideration, though…manipulating the audience’s attention is a very important artistic part of being a director. At the micro level, there were some shots and transitions that felt a little off, like Jones was in a hurry to get to the next scene, or like he really enjoyed this scene and wanted to waste time on it just cause it was pretty, or clever, or whatever.

Anyway. I guess my overall verdict is “good concept, but just okay execution”. One of my friends said “the beginning and the end were good, but the middle wasn’t,” and I can totally sympathize with that. I think that’s because the beginning and end were Jones actually getting to play with the premise, while the middle was just taking care of that annoying plot. And unfortunately it showed.

It might be more honest than Inception, but it came off with less emotion. It’s a lot better than The Adjustment Bureau, though (another concept-driven movie, but one that totally fails to build on its concept). Was it worth $10.50, plus $6 for popcorn? No, that’s ridiculously overpriced for popcorn. I’m not sure about the $10.

$10 for an interesting premise, some nice special effect shots, and a story you haven’t seen before. Well, that last is pretty valuable. But I can’t see myself ever getting the urge to watch this again in a couple years.

The rest of this post is about the ending(s). Stop now if you don’t want spoilers, though this whole section assumes you’ve seen the movie.

Our group came up with three places the movie could have ended; I came up with two more today. Here’s how I think of them:

  1. The Christopher Nolan: End with the same sweeping shot of the duck, the lake, and the train, when Goodwin sends Colter back into the source code for the last time. Up in the air if he actually manages to save the train, if he experiences life in the source code universe, if he dies nonetheless.
  2. The Indie Film: End with the freeze-frame. His life ends frozen in a perfect moment, and in the end that’s what’s important. He has done his duty, and Goodwin does the honorable thing.
  3. The “How Avatar Should Have Ended”: Colter only gets eight minutes, because that’s all Sean had. He saves the train, gets his perfect moment, but dies at the eight-minute mark anyway.
  4. The Fairy Tale: End with them happy in Chicago, never show the CAOCN facility again. Then it’s about the romance, not the premise or the science. (This is the lamest ending, in my opinion. It’s also basically how The Adjustment Bureau ended.)
  5. The Duncan Jones: End with the CAOCN facility (i.e. this is the real ending). Explain what happened, in terms of the science and the premise.

In hindsight, the ending Jones chose makes a lot of sense, and explains a lot of what went on in the movie. See, it was weird enough that Colter could leave the train and go places that Sean hadn’t. But it was downright stupid that Colter kept insisting that the source-code-world was real. It’s a simulation. The fact that they could keep running it over and over sort of proved it…you can’t create infinity in a box. It was presented in the way that a sci-fi person has learned to accept as a condition of the world, a rule about the magic system. I seriously wanted to slap Colter or something for not getting it…then it turns out he’s right all along.

In hindsight, this is why Jones wouldn’t let it go. He knew Colter was right, so it wasn’t weird that he was so fixated on this fake world. But if he just presented it as a psychological twitch (he’d rather live in that world than in his capsule, especially when he found out he was essentially dead), it would have worked. Even with the same ending: they really are creating new universes, universes in which “Sean” makes different choices each time.

And I really appreciated the cyclicity (probably not a word). That was more clever than resetting time, really. But it felt like a really tortured path to get there. The rises and falls in the movie were thrown off because of the premise.

But for any of the other endings to work, you’d be leaving out a bit of the premise. Guess I don’t have a solution.

I’m going to finish it off with two interesting questions we thought about last night:

  • What happens to the original Sean-mind in the new universe?
  • What if Sean had been the bomber?

It’s questions like this that make sci-fi interesting! So…while I’m not sure Jones pulled off his movie, I’m really glad to see sci-fi that doesn’t use stock stories. Despite not really liking the movie in the end, I’m hoping directors like Jones don’t give up.

P.S. At the end of the movie, people were leaving the theater. One person called out, “Excuse me, you dropped this!” and held up an umbrella. It was awesome and the whole theater laughed.