Most time travel movies rest on deeply confused models of past-future-past causation. So let's take a moment to appreciate the gems that don't. You can, I promise, learn to love causal loops.
Just one more satisfying example: In Season 6 Game of Thrones establishes logically consistent time travel via Bran's psychic power to influence the past rather than physically travel to it for this purpose. They use it as a backstory explanation for Hodor - future Bran accidentally traumatized a young Hodor with a vision of adult Hodor's death. Whatever GoT's other sins may be, they got that bit right.
I still can’t understand how the timeline in the original Terminator makes sense.
The idea of an unchangeable future being shaped by the actions of ‘ppl of that future’ who have traveled backwards in time makes sense as a straightforward formulation, but I find it difficult to grasp when I look at concrete examples. I find it especially perplexing to consider that one’s own existence relies on their future self committing some action, I.e. John sending Kyle back in time to save his mother and help conceive John himself. Is it not a contradiction what John must exist in order to ensure his very existence?
I answered a similar sort of question in a post at https://scifi.stackexchange.com/a/78864/22250 with a thought-experiment about how you could generate simulations of self-consistent time loopy universes if you had unlimited computing power--basically the idea is that you don't generate the simulations dynamically, taking some starting conditions and running them forward, but instead generate a huge number of simulated universes with random patterns of events, and then throw out all the ones that don't respect the right laws of physics at every point.
This actually has some parallels to the way Einstein's theory of general relativity works--the fundamental equations aren't dynamic ones that give you later conditions from initial conditions, rather they're just a set of localized constraints that have to be respected at every point in an entire spacetime in order for it to qualify as a "solution" to the theory. If we imagine a God's-eye view of all mathematically possible complete spacetimes with matter/energy fields defined on them (akin to my imaginary computer looking at a huge number of random patterns of simulated events), valid solutions to Einstein's equations would be a subset of these, with the rest violating the equations at some or all points.
It also turns out that there are valid solutions in general relativity which include objects that loop around through time and run into their former selves, their paths through spacetime being called "closed timelike curves" (Kurt Godel discovered one such solution involving a rotating universe, and later Kip Thorne discovered a more localized one involving a traversable wormhole). But physicists think general relativity will turn out to be an approximation to a future theory of quantum gravity, and it's an open question whether quantum gravity will allow closed timelike curves, with some theoretical arguments that the "chronology protection conjecture" proposed by Stephen Hawking will be built into the theory.
Leaving aside rather simple problems such as the continuing existence of the past (if it does then we are its future, so why are we not the past of someone else's future, meaning the end already exists, does this mean everything is predetermined?), whether existence is continuous or discreet, if discreet could someone going back end up between two stages?, how long does it take to reinsert oneself into the past and so on.
The biggest problem I see is that of astrogation and transport. Well's Time machine, The Terminator, or Heinlein's The door into the summer etc. seem to operate on a pre-Copernican assumption that the earth is the center of the universe that revolves around it.
More realistically, our solar system orbits the galaxy at ca. 220-230km/sec, about 1 LY every 1300 years. Our galaxy is moving towards the Great Attractor at +/- 600km/s and so on. Anyone who wants to go back has to calculate the exact landing position, needs to know everything that has happened between now and then that may have disturbed the orbit&position of the solar system and earth's around our sun... Once someone has calculated the position, he/she must transport the traveler more or less instantly without killing him/her either through acceleration or bumping into stuff. But nothing is really instantly. So, when the traveler gets inserted into "past reality", does it matter that this will take time? Head first, feet first?
My point then is, what can we learn from something that cannot happen (based on the thoughts above)?
Just one more satisfying example: In Season 6 Game of Thrones establishes logically consistent time travel via Bran's psychic power to influence the past rather than physically travel to it for this purpose. They use it as a backstory explanation for Hodor - future Bran accidentally traumatized a young Hodor with a vision of adult Hodor's death. Whatever GoT's other sins may be, they got that bit right.
I still can’t understand how the timeline in the original Terminator makes sense.
The idea of an unchangeable future being shaped by the actions of ‘ppl of that future’ who have traveled backwards in time makes sense as a straightforward formulation, but I find it difficult to grasp when I look at concrete examples. I find it especially perplexing to consider that one’s own existence relies on their future self committing some action, I.e. John sending Kyle back in time to save his mother and help conceive John himself. Is it not a contradiction what John must exist in order to ensure his very existence?
I answered a similar sort of question in a post at https://scifi.stackexchange.com/a/78864/22250 with a thought-experiment about how you could generate simulations of self-consistent time loopy universes if you had unlimited computing power--basically the idea is that you don't generate the simulations dynamically, taking some starting conditions and running them forward, but instead generate a huge number of simulated universes with random patterns of events, and then throw out all the ones that don't respect the right laws of physics at every point.
This actually has some parallels to the way Einstein's theory of general relativity works--the fundamental equations aren't dynamic ones that give you later conditions from initial conditions, rather they're just a set of localized constraints that have to be respected at every point in an entire spacetime in order for it to qualify as a "solution" to the theory. If we imagine a God's-eye view of all mathematically possible complete spacetimes with matter/energy fields defined on them (akin to my imaginary computer looking at a huge number of random patterns of simulated events), valid solutions to Einstein's equations would be a subset of these, with the rest violating the equations at some or all points.
It also turns out that there are valid solutions in general relativity which include objects that loop around through time and run into their former selves, their paths through spacetime being called "closed timelike curves" (Kurt Godel discovered one such solution involving a rotating universe, and later Kip Thorne discovered a more localized one involving a traversable wormhole). But physicists think general relativity will turn out to be an approximation to a future theory of quantum gravity, and it's an open question whether quantum gravity will allow closed timelike curves, with some theoretical arguments that the "chronology protection conjecture" proposed by Stephen Hawking will be built into the theory.
Leaving aside rather simple problems such as the continuing existence of the past (if it does then we are its future, so why are we not the past of someone else's future, meaning the end already exists, does this mean everything is predetermined?), whether existence is continuous or discreet, if discreet could someone going back end up between two stages?, how long does it take to reinsert oneself into the past and so on.
The biggest problem I see is that of astrogation and transport. Well's Time machine, The Terminator, or Heinlein's The door into the summer etc. seem to operate on a pre-Copernican assumption that the earth is the center of the universe that revolves around it.
More realistically, our solar system orbits the galaxy at ca. 220-230km/sec, about 1 LY every 1300 years. Our galaxy is moving towards the Great Attractor at +/- 600km/s and so on. Anyone who wants to go back has to calculate the exact landing position, needs to know everything that has happened between now and then that may have disturbed the orbit&position of the solar system and earth's around our sun... Once someone has calculated the position, he/she must transport the traveler more or less instantly without killing him/her either through acceleration or bumping into stuff. But nothing is really instantly. So, when the traveler gets inserted into "past reality", does it matter that this will take time? Head first, feet first?
My point then is, what can we learn from something that cannot happen (based on the thoughts above)?