Back to Back to the Future

My Back to the Future post got quite a response. I’m going to try to respond here to some of the questions/criticisms it received.

Why are you so worked up about this?

Am I worked up? I don’t feel worked up. Unless by “so worked up about” you mean “really interested in.” I’m interested in fictional depictions of time travel because I’m a science fiction writer. In fact, I’m working on a book right now, tentatively titled Schrödinger’s Gat, which deals with some of the conundrums inherent in time travel, and I’m keenly interested in making sure the story holds together, both logically and scientifically. So I like to look at other time travel stories and see if and where they go wrong, so I can avoid that problem. You may think that’s silly, but if I listened to people that said I was being silly, I wouldn’t be writing books in the first place.

Why are you trying to wreck a perfectly good movie?

I’m not. I’m pointing out how it was already broken. If its being broken in that way doesn’t bother you, that’s totally cool with me. And you know you don’t have to read this, right?

Of course Back to the Future is impossible. All time travel stories are impossible.

OK, first we need to determine what you mean by impossible. If you mean physically impossible, I would argue, first, that there isn’t anything in the known laws of physics that prohibits time travel. It may be extremely difficult or infeasible, but it’s not impossible, as far as we know. Second, physical impossibility isn’t as big a deal to me as logical impossibility. For example, it’s physically impossible for X-Wing fighters to bank against empty space while making cool screeching sounds. But I allow that violation of the laws of physics because dogfights in space are awesome. Nor does it bother me that in the movie Up, a man pilots a house suspended by thousands of balloons to South America. I just figure that the laws of physics in the Up universe work a little differently than here. No big deal.

Logical impossibility is a whole other thing. Something that’s logically impossible can’t happen in this universe or any other universe (unless that universe’s laws are so bizarre as to be completely incomprehensible). For example, it’s logically impossible that both of these statements are true:

(p1) At point in time t, Obi Wan Kenobi has met C-3PO

(p2) At point in time t, Obi Wan Kenobi has not met C-3PO.

A logical impossibility is a type of plot hole, and as such it’s a failure on the part of the storyteller.

But time travel is inherently illogical: If I go back in time and kill my grandfather before I’m born, then I never existed, so I can’t go back in time to kill my grandfather. It’s self-contradictory.

It would be self-contradictory if you could do that. In fact, that’s the exact sort of contradiction that occurs in Back to the Future: Marty goes back in time and changes history, so that the Marty that went back in time never existed (although a different but similar Marty does exist). But you’re assuming that time travel allows that sort of impossibility. Maybe it doesn’t. There are lots of sci-fi stories, in fact, where the protagonist goes back in time to alter history and fails. In some cases, as with Philip Dick’s story “The Skull,” the protagonist actually inadvertently brings about the very event he is trying to prevent. The whole plot of Twelve Monkeys revolves around the fact that although time travel is possible, changing history is not. In The Terminator, John Connor sends his friend Reese back in time to stop his mother from being killed, and Reese ends up becoming John’s father. History ends up happening exactly as it did “the first time around.” If history resists being meddled with (or, more accurately, the effects of any travel to the past have already come about), then there’s no inherent logical contradiction.

But how can John Connor send Reese back in time to become his father? That means Reese being sent back in time is dependent on John Connor existing, and John Connor existing is dependent on Reese being sent back in time. It’s circular!

Yes, it is. But circularity doesn’t imply logical impossibility. It’s counterintuitive that A can cause B which causes A which causes B, but there’s no logical problem with that relationship. You were probably taught at some point that circular reasoning is a bad thing because it doesn’t get you anywhere: there’s no way either into or out of the circle, so it’s a pointless form of argument. But there’s nothing inherently wrong with circular reasoning. It’s perfectly acceptable to say that 1 + 1 = 2 because 2 is defined as the sum of 1 and 1. It’s somewhat pointless, because you’re just saying the same thing in two different ways, but both statements are perfectly acceptable and true. The relationship is valid; it just isn’t particularly helpful.

In any case, we aren’t talking about circular reasoning: we’re talking about circular causation. A prohibition against circular reasoning doesn’t have anything to do with the possibility of circular causation.

But something has to start the circle going!

Why? Think of it this way: Let’s say that we have a basic premise that says:

(p3) Every event must have a cause.

Does our circular causal relationship violate that principle? No, because A has a cause (B) and B has a cause (A).

OK, so let’s add another premise to try to negate the circular causality:

(p4) The cause of an event must precede the event.

Does our circular causal relationship violate that principle? No, because thanks to the wonders of time travel, we can send someone in the past to do A, thereby causing B, which brings about A. Every cause precedes its effect. Hmm. OK, let’s try adding another premise.

(p5) In addition to the causes within a system, the system must itself have a cause.

This seems pretty solid, but I don’t see any reason that it’s necessarily true. And if it is true, we have another problem: let’s say we postulate a third event, C, which initiates the causal relationship between A and B. First, the presence of C doesn’t do anything to the relationship of A and B. A still causes B which causes A which causes B, etc. The only difference is that we now have an additional cause that sets up the relationship between A and B. And what caused C? Presumably another cause, D. And on and on. So now instead of an infinitely circular causal relationship, we now have an infinite series of causes (which can still hold infinitely circular relationships). I don’t see how that’s an improvement.

OK, so what if we just postulate that:

(p6) For any event, an effect of the event cannot be the cause of the event.

Well, that does it. If we accept (p6), then our circular causation is impossible.

But what reason do we have to believe that (p6) is necessarily true? That is, it certainly appears to be true in my experience, but it seems like a statement of how our universe works rather than a statement of how every universe has to work. In other words, it implies that time travel is physically impossible (in this universe), but not that it is logically impossible. To see the difference, consider the fact that while it’s impossible to imagine a universe where 1 + 1 = 3 (because it’s self-contradictory), one can easily imagine a universe in which an effect of an event can be a cause of the event. For example, a universe in which it’s possible for a person to travel into the past, thereby causing something that has already happened. In fact, I’ve seen many fictional depictions of such a universe, such as the Terminator movies.

Of course you can strengthen (p6) as follows:

(p7) For any event in any universe, an effect of the event cannot be the cause of the event.

But I don’t see how you can possibly know that (p7) is true. Even if we accept the dubious premise (p6), I don’t see how you can know that it’s necessarily true in all possible universes. If I claim that 1 + 1 = 3, you can easily show that this belief leads to all sorts of absurd situations. But if I deny (p7), it has absolutely no effect on reality as I know it. It doesn’t contradict anything else I know to be true. So it’s hard to see how (p7) is logically necessary.

But infinite circularity is inconceivable!

Yep. Unfortunately, inconceivability isn’t a good test for whether something is logically possible. I find a lot of quantum theory pretty inconceivable, and yet it seems to be borne out by experimentation. The size of the universe is inconceivable to me, but it doesn’t seem to feel any obligation to shrink to accommodate my imagination. I find it inconceivable that the universe exists at all, in fact, but there it is. The only way to see if something is logically impossible is to determine whether it contradicts something that is known to be true. As far as I can tell, time travel doesn’t.