How substantial does the time travel have to be? Do humans have to be able to go back in time, or would discovering Planck-scale closed timelike curves count? If the former, this is pretty much guaranteed to resolve N/A, but if it's the latter, there's a small chance that they will be discovered.

@JosephNoonan Short duration can count, even very short, but if it boils down to "we did a thing in a lab and read some numbers off a screen" rather than "I went back in time and talked to my great-grandfather" then there needs to be very strong proof that it's really time travel and not just a lab error or some fluke of quantum physics that we don't fully understand yet.

@A Also something like "technically positrons are electrons travelling backwards in time" or similar would not count as time travel if it doesn't actually change any observable properties of the universe for us.

@A Absorber theory being confirmed (perhaps in some more advanced iteration) wouldn’t count either I guess? What if there is strong evidence in favor of a theory that predicts time travel but also predicts that it cannot be possibly observed? By comparison GR solutions like the Schwartschild metric work perfectly behind the horizon at r=rS but also predict that while it’s possible to go there you won’t be able to come back or signal back. In that sense the region behind the horizon exists (within the theory) but is predicted to be inaccessible empirically.

@mariopasquato From what I can understand of absorber theory it wouldn't cause this question to resolve, though if there's a big breakthrough there maybe we'll have to ask some real physicists their opinions. If we don't observe any new time-travel-like phenomena and it's just new math for explaining existing physics that wouldn't count.

I'd be skeptical about any theory that predicts time travel that can't be observed, since how do you know the theory doesn't break down under the extreme conditions that supposedly cause time travel? I think we would need to at least have some kind of direct observation related to the time travel itself in order to resolve this question.

@A Ok perfect! Thanks

How on earth could it be fixed? What force prevents you from changing the past?

@asmith There's a principle in physics called the Novikov self-consistency principle that says that time travel must always remain self-consistent, i.e., you can't change the past. That's actually a logical principle, though, not a physical one, because there's not a specific force that makes this happen - it's just that any scenario where the past changes is logically inconsistent and therefore can't happen in real life. It would not be a solution to the equations of physics (whether that the field equations of GR, or some theory of quantum gravity), since it would give two contradictory answers for what happens at some time.

If you really do go back to the past, the very same past that you came from rather than a different timeline, then everything you do in the past already happened. It's part of the same timeline you came from. So you can affect the past, but you can't change the past. For example, let's say I go back in time and bury a chest. That means that the chest was always there before I went back in time, since I buried it before I went back in time. It was always there because future me had put it there, so I affected the past by putting a buried chest in the past. But there was never a version of the past where the chest wasn't there, so I never changed the past.

@asmith Fixed timeline is the only kind that has actually been proposed as potentially consistent with general relativity.

The "force" that prevents you from changing the past is the same one that prevents you from changing the present or the future; if things are a certain way, they are that way and not a different way. If things will be a certain way tomorrow, they will be that way tomorrow and not some other way. And if they were some way yesterday, they were that way yesterday and not some other way.

Suppose this sequence of events:

Day 1: you work on your time machine.
Day 2: you invent your time machine.
Day 3: a "copy" of you steps out of the machine and tells you that he is you from the future. He proceeds to do X, Y, and Z, then walks away into the distance.
Day 4: you step into the machine and vanish into the past
Day 5: your "copy" walks back into view from the distance.

What happens to you subjectively when you walk into the machine? You tell your original that you are yourself from the future, and then you do X, Y, and Z. What stops you from doing something different? The same thing that stops you from doing something different from what you are actually doing at any other given time.

You might object: you could precommit to doing Y if the copy does X, Z if the copy does Y, and X if the copy does Z, before you build the machine or enter it.

The answer is that if your precommitment is effective, you will never enter a time machine, because your physical state (i.e. this effective precommitment) was inconsistent with the physical state of entering a time machine.

i think a fixed timeline is ex ante the most likely. but bayes-wise, if we discover time travel, then surely it wouldn't be a fixed timeline, otherwise we should have seen time travelers at some point in history.

@SemioticRivalry Depends on what times it's possible to go back to. If you travel through a wormhole to the past, then the other end of the wormhole must have already been there in the past. But are you able to create that end from the future, or can you only create it in the past? If the latter is the case, you can't go back in time to before time travel was invented, since no one back then could have created the other end.

We're never going to discover time travel, but if we do, I'd give it about a 90% chance of being the fixed timeline variety, and the remaining 10% goes to multiple timelines. The "dynamic timeline" idea is not even logically coherent, so it has a 0% chance of being true in any possible reality.

@JosephNoonan but it makes the best time travel movies

@Stralor anyway, I think this question is more about the nature of reality. are we in a single deterministic universe? is fate a lie? are there other universes like ours?

@Stralor I don't think this market directly answers any of those. Determinism has nothing to do with which models of time travel are possible. A single, fixed timeline or multiple timelines are both compatible with determinism and indeterminism. A single, fixed timeline also doesn't imply fate - that timeline is bound by the laws of physics, including the physical laws that lead from your actions to their consequences. And other Universes could exist without having anything to do with time travel.

@Stralor

but it makes the best time travel movies

The single, fixed timeline is actually my favorite in movies as well, for a variety of reasons. The dynamic timeline only seems to make for the best movies because it's been used so much more often than the other options.

@Stralor Incorrect, precisely because it is incoherent, it cannot make a good story.

The only kind that makes a good story is the fixed timeline type.

@JosephNoonan If anything fixed timeline time travel can be indeterministic even if the laws of physics are 100% deterministic, because there can be causal loops that have no other explanation for why they are that way instead of some other way (e.g. you step out of the time machine, hand yourself a penny that you carry into the time machine, without the penny having any origin)

@PlasmaBallin I don’t exclude though that we may discover something that barely qualifies as time travel and is so alien with respect to our current prevailing concepts of reality that it ends up being described in terms of dynamic timeline metaphors. Take a relational picture where the only thing that physics is allowed to describe is the information that an agent has on its environment. I don’t see much preventing this information from evolving so that the agent believes that its timeline is dynamic. If we drop the distinction between what the agent believes and what is then we may end up with what looks like an open timeline. But this seems very far fetched.