The thing about this "self-correction" is that nobody, but nobody has a mechanism to explain this force. Not how it works or worse yet, how it will know how to self-correct. What particular force decides that the friction on some surface will briefly rise and cause the gun to jam? How does it know to do that?
In all of the paradox-forbidding time travel scenarios, nobody has ever mentioned anything deeper than "Well, uh, reality just won't let that happen." If I right now posit that each time travel scenario spawns an intelligent angel whose sole job it is to dash about a closed timelike curve invisibly nudging away paradoxes through minuscule miracles of improbability, that's quite literally more than you'll see written about in these situations.
I think the flaw here is assuming "knowing" rather than a mechanism that makes it impossible to express the paradox, or alternatively, impossible for you to find yourself in a world where an action that created a paradox happened.
One option: We tend to see time as mutable, but it's not clear that time needs to be mutable. Imagine time as a window into an immutable graph of deterministic state transitions, for example. You could come up with such a model where loops "back in time" were possible, but where the model simply can't express situations that leads to paradoxes, and where we by extension can't exist in a universe where people can ever take steps that would lead to a paradox.
That option implies time travel is not possible, since your mere existence at a past moment is already mutating it. Which is the opposite of what the article claims.
My description explicitly hypothesised a universe that is entirely immutable, and where the passage of time is purely "read only". Your presence at any given point in time then would never mutate anything ever, any more than watching a movie mutates the movie.
EDIT: You can certainly imagine more "relaxed" models that would also work, where mutability is possible, but where constraints increasingly restrain which mutations are possible when looping back on itself, but an immutable model is the simplest to describe in order to illustrate that models that allows loops can be made self-consistent.
Why? If time is immutable (essentially a Block Universe where everything is predetermined) then you traveling in time doesn't mutate anything, just as you experiencing the current time isn't mutating anything.
Of course it (seems to) follow that you don't have any free will. But there's nothing in science or philosophy contradicting that (if anything we seem to find more and more evidence to support that idea).
E.g, a possible example immutable scenario that doesn't violate anything (with each number representing a discrete period of time)
1. An old man appears and says hi to me and disappear.
2. 20 years pass.
3. I obtain a time machine.
4. I travel back in time and return (in my absence from this time period, I'm at time step 1 saying hi to myself)
The key being that we tend to see this as:
1a. Nothing happens.
1b.An old man appears and says hi to me and disappear.
2. 20 years pass.
3. I obtain a time machine.
4. I travel back in time and return (in my absence from this time period, I'm at time step 1b saying hi to myself).
At which point 1b looks like a change from 1a, so we conclude there's mutability. But if the universe is immutable, then 1a never existed in the first place. I always, immutably had will have (sorry, I could not resist) travelled back in time to greet myself
And that removes e.g. the grandfather paradox, because a static, immutable representation of it that is consistent with our other rules (not having people spontaneously come into existence, for example) isn't possible.
It does, but that's also the point. The past is fixed in this model. If you go back in time, that's not new. You always did that. How the present looks now is a function of you having always done that. We know that you didn't shoot yourself, because you didn't.
Yes of course, except for the last part. Their temporal experience, if they went into a time machine, is:
1. Time machine backwards
2. Shoot themselves
3. Time machine forwards
And that’s how they would have experienced it. The whole premise seems to be that this is allowed:
1. Time machine backwards
2. (do something)
3. Time machine forwards
Of course step 2 always happened since it’s the past. And yet they shot themselves. But they couldn’t shoot themself and exist now as well. And that’s the paradox.
Two choices at this point:
1. You can’t really do anything in step 2. In which case time travel is impotent, and hence why even dignify calling it time travel at all?
2. You have some potency in step 2, but you can’t do things that lead to paradoxes. But then how would you handle those special cases?
Just stating that “the past is fixed” doesn’t solve the paradox, it restates one of the premises.
You're still thinking of this as a chain of events unfolding, rather than as a playback of a set of immutable, self-consistent events.
> The whole premise seems to be that this is allowed:
No. That was not the premise. The premise was an immutable representation of self-consistent state transitions, so what occurs in step 2 will always have occurred at that point in time, because otherwise the system isn't immutable. Step 1. just links step two into the causal chain of events in a way necessary to make the existence of 2. consistent.
Again: Think of a graph. An immutable graph. The flow of time in this model simply follows the graph. It doesn't do anything at all, it just replays the events.
The problem here is thinking about this as if there are choices being made as time passes.
> Just stating that “the past is fixed” doesn’t solve the paradox, it restates one of the premises.
It's not stating the past is fixed. It's stating everything is fixed.
And I’ll reiterate: I accept all of these premises! I’ve never denied them! All you and Eli have managed to do at this point is try to nitpick the linguistic deficiencies of my explanations, which are inevitable anyway since I’m not a physicist with a specialization in space-time theories.
And whether I “get it” or not is secondary to the point of whether you have truly managed to confirm or disconfirm the time travel paradox, which I still claim you have failed to do. It is simple to wave away a time travel where you do something compeletely benign, like frickin’ wave or greet your former self. Try dealing with something of consequence, like killing your former self (as in: the person in that point of time-space which carries the mutable signifier of “yourself” which is younger than yourself) or giving yourself a kilogram of gold from the future, which I think would violate the law of preservation of energy or whatever.
Is time travel backwarks possible or not? If it is, see the questions in my previous comment.
> No. That was not the premise.
And yet if time travel backwards is possible you simply cannot deny that one would experience (!) that time travel as steps 1–3. Temporal experience. Hah!
There’s absolutely nothing hard about the concept of time being immutable; that’s how children’s cartoon handle time travel, not as two timelines where the first one gets erased right after the second one “happens”. (Okay, okay, they probably use both of these tactics.) The true puzzle is time travel which introduce paradoxes, which your original comment says absolutely nothing about. But maybe you’ve answered that in another comment (I have only kept up with this subthread). If not, you should probably try to do that instead of dewelling on this minor philosophical insight of yours.
You've made a hell of a mess of all this. The premise presented is simple. The past is fixed, the universe is consistent, if you travel back in time that always happened.
You're saying there's a puzzle about something that the premise at hand simply says is impossible. You might as well imagine being able to create energy out of nothing and then say that's a puzzle in a universe in which we have conservation of energy. Just because you can imagine something impossible doesn't make it a paradox.
And I’ll reiterate: I accept all of these premises
No you don't. You keep saying "but what if you go back and change the past?"
> Try dealing with something of consequence, like killing your former self (as in: the person in that point of time-space which carries the mutable signifier of “yourself” which is younger than yourself) or giving yourself a kilogram of gold from the future, which I think would violate the law of preservation of energy or whatever.
You can make up impossible scenarios without time travel too. Like someone that shoots themselves (lethally, thorough and immediate) and then a month later they go out to lunch and eat a ham sandwich.
The ability to write out a scenario doesn't mean the scenario is possible.
The theory is that "I go back in time and kill myself" is just as impossible as "I kill myself then a month later eat a ham sandwich". And for largely the same reason: both scenarios violate causality, so the universe can't be shaped that way.
Shooting yourself and then later going out to lunch is simply impossible; it is not a paradox.
A paradox is when two contradictory things are true at the same time.
The grandfather problem is a paradox if time travel is possible. If it isn’t possible then it is simply an impossibility.
> And for largely the same reason: both scenarios violate causality, so the universe can't be shaped that way.
So you agree that time travel is impossible. Great... we agree. However I was discussing this with someone who apparently thinks that it is concievable... I wasn’t merely making up impossible scenarios; I was demonstrating how their theory gives rise to paradoxical scenarios.
The key here is that, if time travel is possible, you would experience the events like this :
1. You start your day normally
2. You meet yourself from the future and discuss some niceties.
3. Many years later, you enter a time machine, and you arrive back at time 2, where you have that conversation with yourself that you remember.
The past has already happened, so it can't change. So, if you haven't met a time traveler so far, that is proof that no one will ever time travel to the moment and place where you just were, even if time travel becomes possible in the future.
Now, you may ask, 'what happens if at step 3 I decide to shoot myself, or not have that conversation?'. Here the answer is simple : that question makes no sense. You have already done our not fine these things, there is no power of decision that could influence this. It's like asking "what if I now decide that yesterday I didn't go to the store". The fact that the event at time 2 happened is fixed both in the past and the future. It was in fact fixed at the time of the universe's creation.
Now, you may not accept determinism, or you may believe in free will, and there are good reasons for both (quantum indeterminism, subjective experience, etc.). In either case, I don't think you can believe in time travel, since all of the paradoxes reappear. But if you believe the past and the future are already set in stone, there is no paradox if the two cross each other in certain ways. But it doesn't make sense about what you 'will decide to do', because this is as fixed as what you ate yesterday.
I think that argument is just recapitulating what a paradox is. It doesn't address "what if you try to do something that didn't happen", which is the root question.
No, the argument is saying that the question you are asking makes no sense. You can't "try" to do anything differently, because these things have already happened.
Again, it's just like asking "what if I try to not go to the shop yesterday, what would happen to the bread I'm eating?". Can you decide today to not do something yesterday? No. In the same way, a time traveler can't "choose" to modify the past - it's something that's already happened.
Of course, this does imply that the future is also set in stone, so in fact you can't choose to go the shop or not tomorrow any more than you can choose to have done so yesterday. This is not a paradox, it's just counter to our subjective experience.
Again recapitulating what a paradox is. If time travel works at all (the premise) then I can look up a historical event, and send a hand grenade back through time to that event. What then? The response "you obviously can't do that" isn't answering anything. What happens when I do? Deliberately, to defy the argument that it's impossible. What then? That's the whole issue.
Please answer this question: what happens if, today, you decide to not have gone to the store yesterday? What stops you from deciding this?
The way this particular idea of time travel works is like this:
1. A time traveler throws a grenade, killing Joan of Arc.
2. History happens.
3. The time traveler who killed Joan of Arc is born.
4. The time traveler goes back in time and kills Joan of Arc.
5. We hear of this event.
If we are looking at this when at time 5, do we find it surprising that it happened? Do we need any explanation? We can ask "what would have happened if the time traveler had instead killed Chairman Mao?", and we can get into a discussion on the possible future life of Joan of Arc etc. But since the event has already happened, we don't need a fundamental explanation for "why didn't it go differently", "what stopped the time traveler from doing differently" etc.
If you then accept that the future is just like the past - it is already decided - then the same banality happens if we learn of this event at time 3. We can ask "I wonder what would happen if this child wouldn't one day kill Joan of Arc", but it is simple musing about alternative histories, like asking "I wonder what would have happened if Hitler hadn't killed himself". The fact is that Hitler did kill himself, and that the child will kill Joan of Arc.
If you imagine the future as changeable, or if you imagine that beings can choose what they will do, then the paradox arises. But if you consider the future as set in stone, and if you consider that humans can't chose to do one way or another, than there is no paradox: history COULD have been different, but it just isn't. The same thing that stops you from changing the past also stops you from changing the future.
Note: I am not claiming that this view is "correct". Just that, if you accept it, than time travel really doesn't give rise to any paradox.
Answer: I have a time machine. If that means anything, it means I can go back to yesterday and decide not to go to the store. Otherwise what is it? Some kind of television into the past?
It is a loop between the present and the past. You keep talking though about deciding to do things, and in this paradigm, there is no such thing. I am not 'deciding' to write this post, it is simply a consequence of the evolution of the universe and my life so far. I may feel like I am free to write it or not, but that is only an illusion.
Just as much, if you have a time machine, and go to the past, you will do exactly what you had already done in the past. It is not something you have a choice on, it is simply a consequence of the universe so far. The same would be true if you went to the future - all your future actions were decided the moment you were born, the moment the Earth formed, the moment the universe was formed. There is no choice.
Ah. Then I decide to disbelieve in the deterministic model of the universe proposed. And I cannot be faulted for that, because its a consequence of the evolution of the universe. So we're done here.
> A paradox is when two contradictory things are true at the same time.
> The grandfather problem is a paradox if time travel is possible. If it isn’t possible then it is simply an impossibility.
It's only a paradox if you're making certain assumptions about how time travel works.
You need to admit that these are assumptions.
Consider the airplane on a treadmill paradox. There is a hidden assumption that it's possible to accelerate the treadmill in isolation to match the wheel speed, but the logic fails because the speed of the treadmill is entwined with the speed of the wheels on the plane.
> So you agree that time travel is impossible. Great... we agree.
No, that's not what I said.
> I was demonstrating how their theory gives rise to paradoxical scenarios.
It doesn't.
You're assuming you can choose your actions independently from the time travel. The theory says that every choice is entwined with the fabric of the universe, so that by the time you see that an instance of time travel exists, the critical non-paradox-causing choices are already set in stone.
> A paradox is when two contradictory things are true at the same time.
No. A paradox is when two contradictory things appears to be true at the same time. It simply signifies that our model is flawed or incomplete.
The grandfather problem is a paradox if our model of time travel is one which is impossible. That is the point. It is simply a way of demonstrating why there needs to be constraints that makes our more naive notions of time travel impossible. As such it constrains the set of possible valid models to those in which the grandfather paradox can't be expressed within the constraints of the model.
> I wasn’t merely making up impossible scenarios; I was demonstrating how their theory gives rise to paradoxical scenarios.
You've done no such thing. The paradoxical scenario you imagine can't be expressed in the model I proposed, and so it solves the paradox.
> Try dealing with something of consequence, like killing your former self (as in: the person in that point of time-space which carries the mutable signifier of “yourself” which is younger than yourself) or giving yourself a kilogram of gold from the future, which I think would violate the law of preservation of energy or whatever.
It is entirely irrelevant what it is. The point is that if the system is immutable it is impossible to represent a combination of state transitions that alters an existing state, as such the set of possible space-time combinations would in that case only include those where the time travel does not introduce a paradox that depends on change.
The same way the timeline does not include any states where I fold an elephant, because it's not possible to do so.
You have not presented any paradoxes that can work with an immutable space-time other than perhaps if said space-time introduces divergent timelines at any point of mutation, in which case the paradox can also easily be resolved.
> And yet if time travel backwards is possible you simply cannot deny that one would experience (!) that time travel as steps 1–3. Temporal experience. Hah!
You can experience it that way, but your "do something" step 2 would be deterministic and could not possibly include steps that introduced a paradox.
> There’s absolutely nothing hard about the concept of time being immutable;
You're still thinking in terms of a mutable space-time. I did not talk about time being immutable, but the entirety of space-time. Your "do something" step does nothing that changes anything in an immutable space-time, because everything is fixed from the start. Nothing moves. No action is taken. Everything is unalterable.
Time, and cause and effect, in that scenario is simply an illusion of consciousness at every point of a fixed 4d (or more) canvas that never changes in any way.
> The true puzzle is time travel which introduce paradoxes, which your original comment says absolutely nothing about.
My original comment gave immutability as an example of a possible solution. Time travel with paradoxes is not possible. That is the entire point of a paradox: The problem as stated reveals a problem with the presumed model. The solution is to identify and remove flaws in the model. Removing mutability is one such solution.
> It is entirely irrelevant what it is. The point is that if the system is immutable it is impossible to represent a combination of state transitions that alters an existing state,
Alters? Aha! But you cannot alter that which is immutable! Seems that you still don’t get the concept.
See how silly it is to language police people on such subjets?
> would in that case [only include those] where the time travel does not introduce a paradox that depends on change.
Highlighted the problematic part with brackets. This is just hand-waving. “Only those”? Your theory would have to be able to explain concrete examples like the grandfather problem, or else it doesn’t rise to the level of being a theory that can explain the paradox.
> The same way the timeline does not include any states where I fold an elephant, because it's not possible to do so.
Irrelevant unless there is an actual elephant folding paradox that I’m missing. Impossible things can be explained; paradoxes cannot since they reveal a contradiction in the model.
> You have not presented any paradoxes that can work with an immutable space-time other than perhaps if said space-time introduces divergent timelines at any point of mutation, in which case the paradox can also easily be resolved.
This is the first time (?) you’ve mentioned divergent timelines or hinted at non-just-straight-line-time. And yes, that would probably be perfectly adequate, and is the best answer you’ve given thus far.
> You can experience it that way, but your "do something" step 2 would be deterministic and could not possibly include steps that introduced a paradox.
“Determinism” is either irrelevant or non-sufficient. “Could not possibly” is pure hand-waving as there is nothing which makes sure that you are able to do something benign (like greet your former self) but not do something like kill yourself.
> You're still thinking in terms of a mutable space-time. I did not talk about time being immutable,
Wow, I have to hand it to you. I did mention “space-time” once (although only once) and I utterly failed to be rigorous and spell out the whole package deal, namely space-time. What an utter embarrassment.
Not that it changes my point in any way.
> My original comment gave immutability as an example of a possible solution. Time travel with paradoxes is not possible.
You can’t resolve the Barber Paradox by simply saying that the Universe wouldn’t allow the barber to shave himself. That’s effectively what you’re doing here, only using vaguely hard science terms with a seasoning of “you still don’t get it”.
> See how silly it is to language police people on such subjets?
If that was what I was doing, yes. But it wasn't, and your example is nonsensical because you referenced a statement where I explicitly pointed out that you cannot alter that which is immutable.
> Highlighted the problematic part with brackets. This is just hand-waving. “Only those”? Your theory would have to be able to explain concrete examples like the grandfather problem, or else it doesn’t rise to the level of being a theory that can explain the paradox.
No, that is not a problematic part at all. The grandfather paradox is a statement of a problem. You can not solve it without specifying a set of alterations or constraints on the unconstrained model that creates the problem. That is the point: The existence of the paradox suggests the model it is posed within is not possible. The solution to a paradox is to resolve the flaw in the model, as you yourself point out next:
> Irrelevant unless there is an actual elephant folding paradox that I’m missing. Impossible things can be explained; paradoxes cannot since they reveal a contradiction in the model.
That the reveal a contradiction in the model by pointing out a means by which the understanding of a model would allow an impossible thing is in fact the point. The solution is to identify constraints that makes the impossible thing impossible to represent in a model. Just like we can point out ways in which folding an elephant is impossible.
> “Determinism” is either irrelevant or non-sufficient. “Could not possibly” is pure hand-waving as there is nothing which makes sure that you are able to do something benign (like greet your former self) but not do something like kill yourself.
Immutability means you can not DO anything at all. No change ever happens. The passage of time in such a model is an illusion. You're still hung up on treating this as a series of events that are happening rather than a replay of an immutable state from a set of such states that needs to be possible to represent.
What makes "sure that you are able to do something benign" is simply an inherent constraint of needing to be able to represent graph of the transitions in an immutable way consistent with what we otherwise think we know about cause and effect. There's no other mechanism needed. A whole lot of seemingly benign things would also be impossible to express because it would cause interactions we're unable to account for, but we would never see the effects of that because those things also could never be expressed and so has never happened. The same way we can't express the states required to allow me to fold an elephant within the constraints of the universe we inhabit.
> You can’t resolve the Barber Paradox by simply saying that the Universe wouldn’t allow the barber to shave himself. That’s effectively what you’re doing here, only using vaguely hard science terms with a seasoning of “you still don’t get it”.
The Barber Paradox is an ontological paradox. We can certainly turn it into a space-time paradox by positing that the events in it are actually played out, in which case it is simply a restatement of the grandfather paradox.
And resolving such a paradox by introducing a constraint that ensures the paradoxical state is impossible is exactly how to do it.
I've not done any handwaving at all - I've given a very simple hypothetical model, and explained how it resolves it by making the impossible states impossible to represent.
Let me try one more time. S(t) represents the state at a given time. Each state is immutable, has always existed and will always exist exactly as it is, entirely unchanging in every way. That makes the following set of state transition a representation of the hypothetical example:
S(1) would in this case be a state before you return through time. S(2) the moment where you might be killed. S(3) a moment where you're either dead or alive after you've returned through time to kill yourself. S(4) the moment you travel back in time. There'd of course in a real scenario be additional states in between, but they are irrelevant. S(5) would be the continuation of the timeline for all entities not travelling back in time.
Now, to introduce the paradox as stated, S(2) must involve you being killed. But immutability requires that S(2) has and always will be the same. It does not change when the time travel "happens", so any effect of the time travel must be consistent with what happened before it, in both previous states (S(1) and S(4)) and the next state - S(3). The time travel alters perspective, it does not alter the model. So in drawing up what goes into each of these states, you have a choice: Either you're always dead when transitioning to S(3) or you were never dead when transitioning to S(3). So either no killing happens, or it always happened and always will. In the latter case, the killing can not happen by you travelling back from S(4) to S(2), because in that case you never existed at S(4) and never will exist at S(4) and so future you never existed at S(2) and never will. To allow the paradox to exist, you need to be able to exist at S(4) to travel back in time, and so the only valid set of state transitions that involves you travelling back are ones where you're still existing in S(3) (and so in S(4)).
The paradox is explicitly the notion that we lack a clear understanding of whether there is a rule that prevents us from introducing an instability. Immutability in this case then introduces a rule that makes instability impossible to represent because that instability requires mutability, and so makes it impossible to represent the paradoxical state in question, and so if the universe is immutable, the paradox is resolved.
You can continue to argue this isn't the case, but you would simply be wrong.
[Incidentally, another way of resolving this paradox is a model where the instability is perfectly fine - where the system is mutable and if instabilities like this are introduced, the system will just keep oscillating between states as a "wavefront" of state transitions keep rolling through. This would require a model with a "memory" of past states to allow transitions to get "undone", but it's perfectly possible to define such a model. The only thing that makes this a paradox is that it appears to be at odds with what we think we know about the world, and so it tells us that what we think we know about the world is wrong.]
I guess the mechanism is just a constraint on the possible solutions for the scenario. As an analogy imagine holding a rope in your hand and wiggling it up and down, this way you can make waves of more or less any wavelength travel along the rope. Now fix the far end of the rope somewhere or close it into a loop and you will no longer be able to make waves of arbitrary wavelength but only integer or half-integer fractions of the rope length.
In case of the loop the two ends of the rope are now identified and therefore must have the some displacement at all times which you only get if the wavelength is an integer fraction of the rope length. I haven't thought about this in much detail but I would guess that if you wiggle the rope at one point the wave now goes both ways around the loop and interferes with itself. Similarly for the rope attached at the far end the wave will be reflected and interfere with itself. In the end the result of the interference is probably strong damping of all wavelengths that do not have a suitable wavelength.
On the next level down you could than analysis this in terms of the atoms and the electromagnetic forces between them which in combination give rise to the macroscopic behavior of the rope. And as I said, I guess time travel self correction would be essentially the same thing, some kind of interference effect disallowing certain solutions. Just like in the loop example where you can not wiggle the rope in a way that the two ends of the rope have different displacements, i.e. you can not go back in time to kill your grandfather and have this prevent you from doing so. There is a consistency constraint by the rope being a loop and the rope not allowing discontinuities in it.
How this would work out in a complex system that involves the real world and people trying to kill other people is probably really far beyond human ability to figure out. Maybe just one neuron in you brain would not fire and prevent you from pulling the trigger. Maybe aliens would capture you on the way to your grandfather. Maybe something would get into the way of the bullet and capture or deflect it. Impossible to say.
I always thought the idea was some kind of many worlds style event where whatever universe you find yourself in after the split is guaranteed to be paradox free because something must have happened to have prevented it
Of course there's a mechanism to explain it: determinism.
It's got nothing to do with 'knowing' anything, it's just that we experience life as a vivid playthrough of causes and effects that could never have started or ended any differently.
Now, you might not agree with that explanation, but it can't be precluded, and it solves all of your concerns above.
This explains absolutely nothing. You time travel back and kill your former self. Paradox. How to resolve? Determinism.
That’s just invalid.
“Knowing” here just referred to the magical force that will keep time travel suicide from happening. But without a magical force you’re even more stumped!
You are still thinking as though someone can 'choose' to go back in time and kill themselves. Choice has no bearing on it. Either something happened, or it didn't.
It is still consistent that a cause is needed for an effect, so an effect cannot eliminate a cause.
You're just describing something that can't/doesn't happen. There are plenty of rules that describe things that can't or don't happen in space.
You wouldn't have any concerns with me saying it's impossible for me to go to the bathroom in my own house and when I walk out, I am in your house.
How does the bathroom door magically know which house I am in?
> The thing about this "self-correction" is that nobody, but nobody has a mechanism to explain this force.
I have a hypothesis: it's about stability. First, let us discard the notion of time "happening" from a perspective of the universe as a whole. Now, let us posit that only universes with stable structures can exist the same way that quantum superpositions can only collapse into certain states based on conditions [0].
In a stable time loop, the 4-dimensional-string model of the universe contains a section that loops back on itself. Irrespective of how any 3 dimensional beings in the string might experience it, this never "happened", it is just a part of the string's structure, which is stable.
This seems strange from the perspective of a 3d observer within that universe because they believe they have free will [1], but it is simply how physics works. A universe where the time travel causes instability isn't destroyed or anything like that, it simply can't come into existence at all according to the laws of physics.
However, I will further posit that, from the perspective of the time-travelling observer, it is possible to change history even in a stable universe! Consider the following scenario: You travel back in time to kill your own grandfather and succeed. Somewhat surprisingly you realize you still exist, evade the authorities, invent a new identity, and raise a family. Later, one of your own grandchildren decides to enact the same experiment you did in the "previous" [2] timeline and travels back in time to kill his own grandfather (that's you, in case that wasn't clear), and does so before you have a chance to kill your grandfather. Unbeknownst to yourself, this is part of the "original" timeline which now proceeds as normal with you travelling back in time to conduct an experiment in which you kill your own grandfather. This knotted structure is also stable, as are indefinitely complex variations [3].
Of course, the much simpler solution is that no universes can exist wherein this kind of time travel is even a possibility in the first place, but where's the fun in that?
[0] consider the double-slit experiment, wherein a wave-like pattern is formed even if only one electron occupies the experiment at a time. The electron states destined to land in the gap regions can't exist in the first place.
[1] and they do, from their perspective. Philosophically this is an interesting argument, but it is irrelevant here.
[2] it might be more accurate to say "meta-concurrent"?
[3] this form of time travel is rare in fiction, but I have at least seen a knot-stable time loop represented in Dark.
I like this explanation, but it essentially doesn't solve the paradox. It just shifts it to knot-stability.
The question is: can I create a situation which can not lead to stability?
And if so, what happens? Will it simply not work? Or will every situation lead to the possibility of stability?
Say, I travel back in time at the moment the universe "ends", and from that point onward in the past I destroy all time machines with a certain technology. Then, no one can fix what I did, and no stable loop could exist.
Now it is the same paradox all over again.
I mean, my guess is that such a situation simply can not exist, therefore I can not experience it - I can not live in this universe where this is possible. So, I can only live in the universe where my technology somehow fails exactly when someone travels back in time to correct my loop.
> The question is: can I create a situation which can not lead to stability?
From the perspective of the universe there is no "create" or "destroy" because time is not experienced, it is just one dimension of its existence. The concept of "create" only exists from the perspective of a 3 dimensional being experiencing the time dimension. We believe we can choose the future because we believe it hasn't happened yet, but that's an illusion of perspective. In a sense, this means that there is no free will, but that's a different argument.
So to answer the question: no, because the fundamental nature physics drops the probability that you would choose such an action to 0, the same way electrons have 0 probability of choosing paths that lead to the gap regions in the double slit experiment.
> And if so, what happens? Will it simply not work? Or will every situation lead to the possibility of stability?
From your perspective, every situation will lead to stability in the end, because the universe where it didn't has a probability of 0.
> Say, I travel back in time at the moment the universe "ends", and from that point onward in the past I destroy all time machines with a certain technology. Then, no one can fix what I did, and no stable loop could exist.
> I mean, my guess is that such a situation simply can not exist, therefore I can not experience it - I can not live in this universe where this is possible. So, I can only live in the universe where my technology somehow fails exactly when someone travels back in time to correct my loop.
Pretty much yes. Though universes in which the situation you describe is impossible to achieve regardless (either because time travel does not exist or because there's nothing you could do to prevent it from being invented again) are significantly more likely.
> The thing about this "self-correction" is that nobody, but nobody has a mechanism to explain this force. Not how it works or worse yet, how it will know how to self-correct.
You could think of it that the universe where we exist must have a stable/consistent state.
If we go back in time and change the state, then that universe can't (didn't) exist, so it can't be the universe that we are in. From our point of view in the present (before time-travelling), the state change was in the past, it will have already happened before we start time-travelling.
This doesn't technically preclude us from going back in time. It's more that the present state (at least as we observe it) is pre-determined, since we already exist in the present.
This. As far as we're aware, physics doesn't understand narrative, and doesn't see any difference between killing a person, moving a gun or even disturbing a speck of dust: all entail things being in the wrong place compared with the original timeline.
You might expect some similarities on a macro scale regardless: perhaps if you kill baby Hitler, Germany can still be expected to end up with a different right wing leader who invades Poland, perhaps even assisted by people who were prominent Nazis in our time, and maybe even loses to the same allies for the same reasons. But on a micro level the differences are huge: the dates will be different, people dying will be different, and the rest of the twentieth century will be different. A century of different atoms bumping into different atoms, different sperm are getting to the front of different queues, and different people being influenced into different decisions results in a different world. Our would-be time traveller isn't even born, never mind motivated to look for Ms Schiklgruber's illegitimate child.
That force might not be necessary. There are at least two ways this would work:
* Events after the time travel event in the past bring about the future from when the time traveler comes. It seems forced, I admit, to assume that it just works out. I can see it working if the time traveler does not exactly know what they did in the past, which could make the paradoxes more beningn.
* If you travel in time, the timeline gets forked at the moment where you intrude. This is what happens in Dragonball Z.
In all of the paradox-forbidding time travel scenarios, nobody has ever mentioned anything deeper than "Well, uh, reality just won't let that happen." If I right now posit that each time travel scenario spawns an intelligent angel whose sole job it is to dash about a closed timelike curve invisibly nudging away paradoxes through minuscule miracles of improbability, that's quite literally more than you'll see written about in these situations.