Basic time travel paradox

Any time travel story about going back to meddle with time are inherently paradoxical. I need help resolving this.

Here's an example of the problem for discussion:

1. Roberta is born.
2. Roberta becomes a lethal assassin.
3. Roberta assassinates Samir.
4. Time travel is invented.
5. Chang wishes Samir were still alive.
6. Chang goes back in time to stop Samir's assassination.
...2.5' Chang kills Roberta.
...3' Roberta does not assassinate Samir.
...3.5' Chang returns to his normal timeline.
7. Hey, Samir is still alive!

Ignoring that Chang miraculously remembers both timelines, and ignoring possibilities about a self-correcting universe that prevents Chang from killing Roberta or kills Samir by other means, you still have this problem:

5' Samir is still alive.
6' Chang has no reason to go back in time to stop Samir's assassination.

That is, going back in time to change it necessarily destroys your motivation for going back in time to change it, and is thus paradoxical.

How can I get around this for a game about time travel?
«1

Comments

  • My only idea is that the process of traveling through time creates a sort of event "stickiness" around the people who are traveling. That is, those folks who took actions toward the change are not affected by the "time amnesia" (timenesia?) that the change implies.
  • The standard answer from a physics standpoint is that you don't go back in time within your own universe: you actually go to an alternative universe that is like yours in the past, therefore you can make 'changes' there because they are merely part of the present of that timeline.

    The more interesting story-oriented answer (used in sources such as The End of Eternity by Isaac Asimov, Doctor Who and other fictions) is that the agent of change is somehow outside time: that is, as a result of travelling into the past and/or changing it, or as a requirement of being able to do so, the agent exists in their own timestream, or a metatimestream, which is immune to the changes they make to the common timestream.

    A more radical answer might be to say "Sure, you can change the past that way: now the version of you who did that no longer exists," or "Now there are two versions of you running around, with different pasts." In other words, confronting the problem head on, rather than looking for a workaround, probably provides much more interesting story opportunities for players.
  • edited August 2015
    There are multiple models available. You need to create the epistemology of time in your world before you can answer questions about how it works in your world. Here is a very simplified but generally useful graphic...

    image

    ETA: According to Einstein's theories, not only can you not change the past, but you also cannot change the future! It has already happened - just as much as the past has already happened - but we just can't see it because of our position, vector, and speed.

  • I'm pretty familiar with all the standard answers, having immersed myself in the literature for a while now. I need a solution where you actually change stuff in your own universe; otherwise, there's no reason to go back and change stuff, right?

    The "agent of time is somehow outside time" doesn't feel more interesting to me, but I want to talk about it to make sure I understand the possibilities.

    In my game setting, AIs are sending PCs into the past to change it for a Specific Purpose. are the AIs and the PCs all "somehow outside time"? If they are, why do they care about changing time? What is their motive?

    I'm not a big Doctor Who fan, so forgive me here. Isn't the show pretty hand-wavy about the how and why? (But not the Who, haha!) If I have to, I can hand-wave it all away, too, but I'd rather wave my hands as little as possible here.
  • edited August 2015
    /nods. Well, I wanted some flexibility when I was writing DayTrippers, and I wanted time travel in both directions. So I waved my hands a bit. What I decided was basically this: #1 if you're careful and #3 if you're sloppy. i.e.: If you're very careful and cause no major changes to occur, the "tendency" of the timeline is to resolve back the way it was (the mysterious stranger is remembered as a dream rather than a literal event, minor changes due to time visitations are explained away as coincidences or "magic", or simply suppressed/forgotten, etc). But if you cause any major changes to occur, you have split the timeline, creating a new branch - from which it will be much harder to find your way "home".

    It may not precisely fit any of the predicted quantum physical models, but it gives me a nice amount of narrative flexibility as a GM, and makes the game more dependent on Player skills and intentions, which is what I wanted.

  • My model is that it takes energy to make changes to the timeline, and that the universe will take the path of least energy to resolve things. And that resolving paradoxes take infinite energy. So that's the conundrum for me, if basic time meddling is a paradox.

    I definitely don't want to split into two timelines. I'm fine with meddling being constrained to small, iterative things--make a dozen trips to subtly change your future--as long as those trips (in a Traveller SF setting) are fun for players.
  • edited August 2015
    In The End of Eternity, the 'Eternals' have their own plan for humanity and want history to conform the those ends, hence their agents pop into trivial points in human history to make changes that will have important, long-term consequences (an early example of the Butterfly Effect theory of chaotic change)

    Charles Stross explores similar ground in his Palimpsest
    novella, though the time travel is dealt with in a grittier way, that faces up to the headaches caused by being part of an organisation that can change history at will. Again, their goal is to steer human history along the course that suits them, rather than letting it happen willy-nilly.

    In both cases, the issue is one of control vs. freedom: the organisations making the changes feel that it is better to have someone in charge, overseeing the big picture for the 'good' of all humanity, but in both cases there is a revolutionary movement that would rather let humanity makes its own choices, even if those are sometimes bad ones. I can quite understand some AIs wanting to make our history 'neater' and less messy, using human agents to make the changes, even when those same agents don't truly understand the significance of the changes they are making.
  • In my setting, AIs can't survive the jump through time and thus need human agents. They also don't have a perfect picture of the past, only vague historical accounts and socio-mathematical models theorizing what a change might accomplish, so they need to rely on their human "boots on the ground."

    How does Stross handle what I am calling the Basic Time Travel Paradox in Palimpsest? So the shadow organization goes to steer humanity in a certain way for their own dark purposes, but each time they succeed, they lose the motivation to start the process of change. Paradox.

    And yes, I'm also aware of the implications of time travel on free will / determinism. Another problem to deal with.
  • I've put some little thought into this. I also recommend M.J. Young's time travel website, it's crazy (like a fox, and just a bit like a manic fox) good and does a good job arguing about the various ways one could think about this.

    The basic thought experiments off the top of my head are as follows, I'll list these merely for a basic overview of the field of thought:
    a) Fixed time: time cannot be changed, all time loops (if any) are already in existence and causality always works. Characters in the story can only act in ways that do not cause paradox, because paradox would only occur if they acted in ways they hadn't acted before (causally speaking).
    b) Many worlds: time travel does not exist, but you may think it does, because you are capable of accessing worlds that look like the past of your own world.
    c) Uncaused time travel: there is only a single time-line, but travel backwards destroys the future, and all events in the time-line do not have material causes; a time-travel event does not have an existing cause in the time-line, because that cause was destroyed with the future the time travel happened from. You can only infer the causes of any time travelers popping up, not directly witness them.
    d) Phantasmagoria option: causing a paradox is possible, but funky shit happens if you do and anybody notices/cares: the world ends or space gods surface or whatnot, and you can therefore pretty much ignore the logic fault, because you've got bigger things to worry about. M.J. Young's clever model goes in this box, basically, by stating that the world ends if paradox occurs.
    e) Hollywood time travel: time travel works if you never look at it too vigorously, with technobabble options to bypass rationality filters in people's brains. This is basically the default assumption in time travel stories. As it says in the tin, it basically works off not thinking too hard about the principles.

    I've used (c) and (d) for gaming myself. Fixed time works with the right rules set (basically you need something that makes it really clear what is True, and explicitly engages the players in cooperation to ensure that no True thing is ever contradicted), Many Worlds is trivial (but not real time travel), and Hollywood time travel is basically just begging the question.

    For practical hands-on gaming where you expect the players to kick the tires, so to speak, my recommendation is option (c), uncaused time travel; I've fiddled with it in e.g. D&D context, which very much needs a clear system without arbitrariness to make it actually gameable, and I don't really see any immediate issues aside from the ability of a time-traveler to basically rewrite reality in very unexpected ways (although not necessarily in a way they can control). It's a very powerful thing to be capable of, but then that's an obvious reason to have it come to play in high-level D&D.

    My D&D scenario for this involves what amounts to a transcendent Heavenly Plane which stands outside time; time (the universe, really) itself is a concrete artifact, a nigh-infinitely long silver thread running through this plane. Touching it at any given point moves you inside time at that point, rewriting the entire history from that point forward with you in it (and removing everybody who entered time after that point from history entirely, if you think about it), with thoroughly normal causality barring the one event of you appearing from thin air into the time-line. At the same time (in Heaven time) as you disappear inside time, everybody who ever steps outside time in the new history appears in Heaven next to the physical timeline thread. (Every time traveler appears instantaneously because from the viewpoint of Heaven time their decisions to time travel occurred at the same instant, when the time-line reconfigured itself; they merely stepped out of time at different points along its length.)

    This scenario has the practical advantage of having real time travel and not having paradox. Aside from the practical disadvantages of being very unfair and so on, it has the logical disadvantage of having "uncaused" events in the world: there is never any causal reason discernible for a time traveler popping up in the world, as the time traveler has always erased their own departure by their arrival (assuming we're speaking about travel to the past, future travel being logically trivial). This doesn't bother me unduely in the D&D context, it's not like the game isn't full of unexplained mysteries anyway. Better than paradox, except for the brain-twisting horror factor of paradox, of course.
  • (Guys, these are great responses. If I seem argumentative, it's because I am challenging ideas, not because I don't love your ideas.)

    Eero,

    I might be daft, but I don't see how most of these actually prevent the Basic Time Travel Paradox. Type-a doesn't address the solution since you can't change time and my premise requires you to be able to change the past.

    b) Many worlds: time travel does not exist, but you may think it does, because you are capable of accessing worlds that look like the past of your own world.
    So if you return "home," you don't actually want to return to where-and-when you left, because that place-and-time is Wrong and you want the Corrected place-and-time, which is actually a different universe altogether.

    Referring back to my example, Chang goes back to 2.5, really 2.5', changes the past, then returns to 7, really 7'. If 7' existed the entire time, why not just go there directly and skip the long trip into the past? Sure, handwave that you can only get to alternate present times by visiting the past and messing with things, but why?

    I find the whole Many Worlds solution to be fairly nihilistic and boring, and even more problematic regarding determinism than usual. ;)

    c) Uncaused time travel: there is only a single time-line, but travel backwards destroys the future, and all events in the time-line do not have material causes; a time-travel event does not have an existing cause in the time-line, because that cause was destroyed with the future the time travel happened from. You can only infer the causes of any time travelers popping up, not directly witness them.
    I interpret this in two different ways:

    c.1.) Uncaused time travel with amnesia: the time travelers, upon changing time, suddenly cease to exist in the past, and their future (present) selves have no motivation to change the past (and so do not travel to the past).

    c.2.) Uncaused time travel that looks a lot like Many Worlds: the time travelers in this timeline don't remember the time travelers in the other one, but they all exist, right? This has all the problems as Many Worlds.

    In either case, don't you have characters who shouldn't remember that they Did The Thing? As soon as they remember, they're back in the old timeline, not the new one.

    d) Phantasmagoria option: causing a paradox is possible, but funky shit happens if you do and anybody notices/cares: the world ends or space gods surface or whatnot, and you can therefore pretty much ignore the logic fault, because you've got bigger things to worry about. M.J. Young's clever model goes in this box, basically, by stating that the world ends if paradox occurs.
    Merely changing the past is inherently paradoxical, as I explained. Are you saying that, as soon as someone tries to change the past in an obvious way, BOOM, the universe ends?

    Not useful to me, nor does it really solve my problem.

    e) Hollywood time travel: time travel works if you never look at it too vigorously, with technobabble options to bypass rationality filters in people's brains. This is basically the default assumption in time travel stories. As it says in the tin, it basically works off not thinking too hard about the principles.
    Sure. We can basically not try to solve the problem, but that's not what this thread is about, right? More <=)
  • edited August 2015
    My favorite is the "two versions of you" model. Have you seen Primer? I'd love to play a game using that model.

    Basically, from point 2.5 to point 6, there are two Changs, but once Chang gets in the time machine at 6, the paradox collapses, because that's the end of the Chang who hasn't popped out of the time machine yet.

    If Chang2 popping out of the time machine and doing stuff would subvert Chang1's reason for entering the time machine, then Chang2 just needs to tell Chang1, any time between 2.5 and 6, to get in the time machine at 6.

    The way I see it, Chang2 gets an infinite number of tries to get this right, so the one that sticks is the one that becomes reality. (Any that fail simply cease to exist at point 6 because look: Chang1 doesn't time-jump -> Chang2 disappears along with all changes from 2.5 to 6 -> Samir is again dead, so Chang1 DOES time-jump.)

    Within this model, you can have loops enclosing other loops, such as a 2002 to 1995 loop inside a 2010 to 1990 loop, without fear. I'm not sure if you can have overlapping loops, though, like one time machine going from 2002 to 1995 and another going from 2000 to 1990. I'm not sure if the same inevitability principle would apply.
  • For game purposes, each of the "tries to get it right" could have some sort of cost for the time machine or the time traveler. We don't play through the failed attempts, but we do assess their probability and then roll to see how many it took and how much you've aged or whatever the cost is. We assess their probability by just how much they fucked up the past, with "how much" determined by whatever criteria seems fun for the sort of game you want.
  • edited August 2015
    >> Primer
    Best time-travel movie ever. Difficult model to think in, though, at least for me. Took hours to unravel it. (By means of comparison, I think Inception was easier to comrehend.)
  • edited August 2015
    Time travel as what-would-happen is really interesting to physicists and mathematicians; for science fiction authors who aren't just writing for gee-whiz-isn't-that-a-cool-idea people, the key is the theme of the fiction.

    Look at Back to the Future, the best time travel film. It's a movie about kids connecting with their parents and recapitulating their struggles generation on generation. So even though Marty "should" fade and disappear the second he messes up his parents meeting, it doesn't work like that because if it did, the emotions of the movie would be all wrong. Marty's challenge is to understand his mother and father, their milieu and their emotions. If he does, the universe isn't turned upside down.

    Look at Primer, the second best time travel film. It's a movie about people that fuck themselves and their relationships up by obsessing over getting something perfect instead of dealing with glorious imperfections and randomness. The main characters discover time travel absolutely by accident. Yet the plot is driven by (one) inability to just accept that things might not be fixable. If time travel were able to actually change the past in that movie then the theme and emotions of the film would be all wrong.

    So I would identify what the emotions and theme of your game are before I did anything else.
  • I like time as a river, you can create local disruptions, but the flow then merges and continues. You're more effective in using time travel to gather information and take it to the present and then act on it. You push too hard against the time stream and it will push back. There probably is no forward time travel since the 'present' is riding on the crest of the wave.

    Possible twist, your attempts to radically change the present keep failing because you are in some other time traveler's past.
  • I also like the TV show Seven Days. You can only go back a week so you can't radically change the history of the world. Also fits the river definition if you think of history locked into the banks of time and are not capable of change.
  • Adam, in your original example, in 3.5, have Chang go to his normal timeline a little bit later than they left. So that they "skip over" the decision point of "Do I have any reason whatsoever to go back in time to kill Roberta".

    I love Feng Shui's (have only read the old one) time travel with its plasticky mumbo jumbo time travel system. You need to control feng shui sites to "win in time". so gameable!
  • Feng Shui also uses the "river" view of time. Which is very improbable and teleological and antropocentric but... easy to intuit and game with.
  • Yeah, Feng Shui's time travel system is perfect for its nature, which is that you want cool action movie places to fight. If you can erase a fight, the game is pointless, so you can't really erase what people do in an "important" spot in history.
  • Adam, in your original example, in 3.5, have Chang go to his normal timeline a little bit later than they left. So that they "skip over" the decision point of "Do I have any reason whatsoever to go back in time to kill Roberta"
    Or, return to the decision point exactly and become the new reason. Say "Hey, I'm you from another time line, jump into this machine and kill Roberta!"
  • Adam, in your original example, in 3.5, have Chang go to his normal timeline a little bit later than they left. So that they "skip over" the decision point of "Do I have any reason whatsoever to go back in time to kill Roberta"
    This was exactly the suggestion I was going to make. If the only paradox you're concerned about is one of motivation, just leave the motivation entirely in the character's subjective past. They remember a false past, which is no longer the case. It's weird, but it's not paradoxical?

    Of course with the AI scenario you describe, where the original motivation resides outside of the PC, this looks like a poor solution on the surface -- but I think it actually produces an extremely delightful, dystopian/dysfunctional setup where the AIs actually lose track of why they did things. They may be aware that they are using time travel to fix things, but they have no way of verifying whether it worked other than debriefing their highly unreliable human agents.

    So in this take you have:

    1. The original reason for the time travel (known to the AI pre-event, forgotten post-event.)
    2. The explanation offered to the agent by the AI when they are sent back (known to the agent and never forgotten.)
    3. Whatever the heck the agents actually do.
    4. The explanation/description given to the AI by the agent, post-event (which the AI remembers going forward.)

  • My standard paradox-free model of time travel is that time travel "unwinds" the period of time between arrival and departure; the present of the time travelers (and of the whole universe) is reset to the arrival time, and there is no future to go back to; history takes the slow way from a suddenly changed situation. In particular, time travel might not be discovered at all.
    The time travel arrival event consists of people and things appearing out of thin air, without any special connection to another time or another universe; time travelers just remember an extent of history that won't repeat exactly because it's the history of a world without them.

    In a RPG this would make most long distance time travel an heroic act: do you want to forfeit the whole world and strand yourself in a foreign and probably unpleasant time to have a chance of preventing or altering some event? Short range time travel would be more exploitable, for example to "duplicate" people and objects.

    Time travel of memories only, like in Groundhog Day, would fit the same model: someone in the "past" remembering a counterfactual "future", with an injection of knowledge that shouldn't exist rather than of more concrete objects and characters.

    A possible variant: keeping around the departure timeline in a completely detached parallel universe, as if cutting and planting a branch of the tree of time.
    Restricting further time travel before the time travel departure event originating the alternate universe might be a good idea,
    Adventures of who remains behind in this alternate timeline might be interesting in a RPG if there are interactions with the real timeline where the time travelers went or in order to pursue multiple paths of alternate history.

    Secondary cruel variant: time machines need to be operated from the departure time, by someone and/or something that's left behind. Who wants to be lost into a second-class universe, or even into complete nonexistence apart from the memories of the time travelers?
  • edited August 2015
    Would this work? Ghost timelines. If nothing else, it's a cool phrase.

    After Chang goes back, the timeline is split in two, at the arrival point. There's the original timeline with the departure, and the new, prime timeline that starts with the arrival. As soon as this happens, the original timeline starts to fade. Everything is preserved up to the moment of arrival, since this is the same timeline as the prime. After that, only events leading up to the departure are preserved. Anything not affecting the departure fades away, and at the moment of departure, this timeline ceases to be. Thus, Chang still has all his memories, because any events experienced by Chang will have been preserved in the ghost timeline. There is only one world, not many, and the actions of Chang are not uncaused. However, there are now two Changs. If Chang travels back into the future to point 6', there's no time travel happening there, since that happened in the ghost timeline, which is inaccessible. But there will still be another Chang who knows nothing about Samir dying.

    So this explanation would mean that 1) when you time travel into your own past, you're cloning yourself, and 2) you cannot travel in time to the time when someone time travels, since all departures are in ghost timelines. You can, however, travel back in time to before the arrival point, and thus create ghost timelines within ghost timelines. But I don't think that would screw with causality. I think that if you change things so that Samir's life is never in danger, then Chang will not appear, except in the new ghost timeline, maybe? It's hard to think about this.

    Edited to add: Yes, I think that'll work. If you change things before Samir's death but don't change that event, this new timeline will still get a ghost timeline with Chang travelling back to save Samir, but that would be a new ghost timeline, separate from the one in the new ghost timeline (nested ghost timelines!).
  • Ah, so we have future characters coming to do missions in the past, and want a time travel theory that supports adventurous missioneering. My own take on the matter would probably be to mix just about all possible time travel models into the pot and make it a dangerous source of tension as to which one might be the true and applicable one, moment to moment or in general.

    Like, you could even have player characters with time travel science skills, and they could be backing different predictions about what is or isn't kosher to do, and then the GM would also be pushing for different outcomes case by case. So you have a character who is convinced that time is fixed and the time travelers can only do the things they had already done from the future perspective, and you have another one who thinks that the future doesn't even exist anymore in any meaningful way after the travel back in time, and the truth about the matter is in tension during play itself. Over time you find out with the group how time travel works in this setting, likely settling with some sort of phantasmagoric hollywood deal.

    Aside from that, a few direct answers to Adam in the interest of seeing if we might finangle out some useful insight here.
    I might be daft, but I don't see how most of these actually prevent the Basic Time Travel Paradox. Type-a doesn't address the solution since you can't change time and my premise requires you to be able to change the past.
    This is partly true, type (a) (the fixed timeline model) simply assumes that paradox is impossible, and therefore each new fact you establish about your story/scenario has to be logically congruent with everything that has been established before. Your basic time travel paradox is not an issue, because it can never happen that somebody would remove their own motivation for traveling back to the past.

    However, this does not mean that you cannot change time! It merely means that whatever change you make, it is in fact a change that the future you had already made, although you very likely did not yet know about having made it (or you knew, and did it nevertheless, perhaps in the belief that you "must" act the way you know the future you did). Many time travel movies and such stories pay token allegiance to this type of story logic: every time travel movie where at the end you find out that the time traveler in fact did exactly the things that would bring about the future that caused them to time travel in the first place is discussing this model.

    The big, fat problem that fixed timelines have is that it is easy to imagine a different paradox, namely the paradox of free will, which comes up here: if we assume that people have free will unconstrained by any sort of mystical forces, and if we assume that a person knows what they will supposedly do in the future, what happens if they specifically decide to do otherwise? If the fixed timeline model holds true, then you either cannot act other than you already know you did, which breaks with the premise about free will, or you were mistaken about what you thought you would do, which breaks with the premise that you actually had correct knowledge about your future actions. So it seems that either people involved with time travel are mind-controlled by Fate regardless of their will, or it is impossible to gain reliable knowledge about your own future actions, such that even if you try to act differently you always end up acting exactly the way you always did.

    That's a different kettle of fish compared to the basic time travel paradox you discussed, though, and might well be more acceptable or interesting or whatnot for a given game. No reason to ignore the fixed time model, it is certainly potentially interesting for gaming as well!

    (How would I resolve the free will paradox in practical gaming? Well, the play of the game would have to reliably establish without fail that no matter what a time traveler did, they would always fail to interfere with their own motivations to time travel. For most settings and genres and such I would probably go with simple accidental events causing attempts at paradox to fail, but I guess that Fate coming in to actively prevent you with supernatural means would be doable as well. Note that while having your car always fail when you're trying to cause paradox by driving somewhere might seem supernaturally unlikely, but it actually isn't when you account for the prior probabilities involved in something like this: given the fact that we know that you didn't show up in time to interfere, it is in fact not that unlikely for your car to break down or some other perfectly natural event to occur, so as to explain why you did not make it in time. The only weird part is that the time traveler knows in advance that something will happen to stop them, but hey that's time travel.)

  • b) Many worlds: time travel does not exist, but you may think it does, because you are capable of accessing worlds that look like the past of your own world.
    So if you return "home," you don't actually want to return to where-and-when you left, because that place-and-time is Wrong and you want the Corrected place-and-time, which is actually a different universe altogether.

    Referring back to my example, Chang goes back to 2.5, really 2.5', changes the past, then returns to 7, really 7'. If 7' existed the entire time, why not just go there directly and skip the long trip into the past? Sure, handwave that you can only get to alternate present times by visiting the past and messing with things, but why?

    I find the whole Many Worlds solution to be fairly nihilistic and boring, and even more problematic regarding determinism than usual. ;)
    You are entirely correct in that there is no logical reason why one couldn't move directly into the alternate world where events have played out the way you wanted them to. However, the cosmology of the many world multiverse might cause interesting limitations, as might the technology used to effect the "time travel".

    For example, it may be the case that your time travel machinery, for whatever quantum babble reasons, is only able to access alternate worlds that are exactly like your own world at some point in its past. Your machine may not be capable of travel "forward" in time in any special manner at all, too. You may, however, have another machine that enables you to put yourself in suspended animation of some convenient sort, or shunt you outside the cosmos to some timeless realm, so as to allow time to fly forwards without you.

    A time travel setting with a setup like this would in fact be entirely satisfying for many people's time traveling fiction needs: you can go back in time to make changes in the past, you can go forward in time to see what changes your actions caused to the world, you can go back again if you want, and so on. The only way to get to your preferred world is in fact to travel to a "past" and manipulate it to, over time, bring about the world of the future you wanted. Paradox is not an issue because the cause of you appearing in the past is not anything that happens in the future of this time-line, it's something that happened in the past in the original universe you came from.

    The obvious downer for dramatic purposes, the "nihilism" you mention, is that it is easy for people to feel that time travel shenanigans are pretty meaningless if you live in a multiverse of infinite possibility; the only thing you gain by traveling to your preferred world is that you get to live in that world, not that you heroically destroyed your own world and made it anew into the world you wanted.

    And, as I acknowledged right off the bat, the many worlds model is not real time travel for most purposes, just something that looks superficially very similar.

  • c) Uncaused time travel: there is only a single time-line, but travel backwards destroys the future, and all events in the time-line do not have material causes; a time-travel event does not have an existing cause in the time-line, because that cause was destroyed with the future the time travel happened from. You can only infer the causes of any time travelers popping up, not directly witness them.
    I interpret this in two different ways:

    c.1.) Uncaused time travel with amnesia: the time travelers, upon changing time, suddenly cease to exist in the past, and their future (present) selves have no motivation to change the past (and so do not travel to the past).

    c.2.) Uncaused time travel that looks a lot like Many Worlds: the time travelers in this timeline don't remember the time travelers in the other one, but they all exist, right? This has all the problems as Many Worlds.

    In either case, don't you have characters who shouldn't remember that they Did The Thing? As soon as they remember, they're back in the old timeline, not the new one.
    I don't think I explained myself very well. The uncaused time travel model does not feature any extra-causal "magical" events aside from time travelers (or time traveling information, in a story with future-predicting magic powers) popping up seemingly from nothing. In general most of these models, aside from the Hollywood one, don't feature magical effects like a person opening a door and therefore forgetting something, or splitting into two different people, or things like that; conservation of credibility and all that, as required of good scifi :D

    This model is very similar to the many worlds model in that it also denies the possibility of paradox by breaking the causal chain between the future and the past. The explanation is different, though: instead of the cause for the time traveler being there and acting in the world existing in the past of another world, the cause did in fact really exist in the "future" of this single universe, except that that future itself got utterly undone by the "rewinding" of time in the time travel act. We are left with an event that truly does not have a mechanical cause in the universe, as the time traveler pops up out of nothingness in his own subjective past, which is why I call this "uncaused time travel" - it is impossible to differentiate between a time travel event and an event originating transcendently, outside the present universe (except by asking the time traveler where they came from, presumably).

    Note that it is likely in many time travel models for the traveler to change the past in a way that causes their future time travel to never occur. For one, they could simply travel far enough in the past to cause a person recognizable as "themself" to not be born in the first place. As much as we know about the biology of human cognition, how weather occurs on the planet and impacts human life, and other such chaotic systems, it would be utterly naive to think that somebody could travel very far to the past and not erase their own future altogether.

    This seeming paradox of a time traveler preventing their own time travel would be a problem for a model of time travel that acknowledges the possibility and necessity of causal chains stretching from the future to the past, such as the fixed time model (a). It is not a problem for e.g. models (b) and (c) because (b) claims that time travel is just shuffling between worlds, and (c) claims that time travel events are, causally speaking, miraculous - they may have had some cause in a world that ceased to exist, but that doesn't matter, and nothing that can happen in the present world may impact that world.

    Uncaused time travel has the advantage of not having to care about paradoxes, which is of course also the disadvantage of not getting to care about paradoxes - it might not feel like time travel if you expect time travel stories to be all about how people cause paradoxes and what happens next.

    d) Phantasmagoria option: causing a paradox is possible, but funky shit happens if you do and anybody notices/cares: the world ends or space gods surface or whatnot, and you can therefore pretty much ignore the logic fault, because you've got bigger things to worry about. M.J. Young's clever model goes in this box, basically, by stating that the world ends if paradox occurs.
    Merely changing the past is inherently paradoxical, as I explained. Are you saying that, as soon as someone tries to change the past in an obvious way, BOOM, the universe ends?

    Not useful to me, nor does it really solve my problem.
    It surprises me a bit that you don't find the phantasmagoric option useful - it is very popular for dramatic time travel stories! The effect doesn't necessarily need to be the immediate cessation of all existence, of course. We might posit, for example, that ugly hunch-backed shadow men pop up out of the paradoxical nether and rip up anybody witnessing the paradox; the truth about reality is that it's just a virtual construct God set up, mechanical causality only exists because God programmed reality that way, and the hunch-backed shadow people are there to make sure nobody notices it when there are glitches in the program. Works swimmingly for certain sorts of purposes.

    As you can see, the phantasmagoric option relies on a trick of psychology combined with the assumption that our world is not real in terms of scientific positivism. The way it handles paradox is not thoroughly logical (for instance, there's no good way to determine when the hunchback shadows come out - when in time does a time paradox occur, exactly?), but it doesn't need to be, because the world is actually a simulation in some sense, and therefore paradoxes may only get caught at certain junctures in the program, when they cause infinite loops or whatnot, just like in literal computer programming; the phantasmagoria, whether shadow men or angels or the world unraveling dramatically starting with the photographs of all your loved ones, may start occurring at whatever moment in the execution of the "reality software" the paradox happens to cause the execution to become buggy.

    Many time travel movies fundamentally imply some sort of phantasmagoric world-view. If I had to describe the standard Hollywood time travel "formula", I'd say that it is a dask of option (a), fixed timeline, and a big portion of option (d), phantasmagoric timeline. This is why so many movies suggest that you cannot change Fate (fixed timeline theme), but attempts at doing just that for some reason seem to cause unexpected and illogical special effects (disappearing mementos, unexplained amnesia, even monsters or the world literally crumbling away street by street) nevertheless, and at the climax of the movie you usually get to see how, somehow, the time travel elves fixed time so that the protagonist got their happy ending and apparently flouted the paradox in some manner.
  • Some background might help.

    I'm creating a setting for the Traveller RPG (working title: Time Traveller), in which a handful of godhead AIs can meddle with time, but only through their human agents, who don't have the same limitations as the software constructs. They're meddling with time to reshape the galaxy for their own purposes.

    Prometheus is trying to evolve humans to their fullest potential. Colonia is single-mindedly committed to exploring the galaxy and establishing human civilization all over it. Utopion is conservative and happiest when people are hibernating in vast virtual realities, and fears the chaotic change of time meddling, but must meddle itself to keep others from changing too much.

    All in all, it's a Traveller game, with starships, wormholes between habitable systems (not jumps per se), some psionics, and most of the trappings of the Traveller RPG. Less empire, though. More medieval in feel, the galaxy just now waking up from a long sleep after a metavirus tore through known space, shutting down travel and most electronics for a hundred years.

    So the ability to tear down the past and rebuild it is core to this setting. The AIs are core benefactors and enemies, so I need them to remember what they changed (vis-a-vis time meddling), and I can hand-wave that to a point, as they're somewhat mystical, unfathomable beings. I envision each of them as type-II Kardashev scale civilizations unto themselves, harnessing a star or star cluster for power and computation. These beings can't access the wormholes that the starships fly through, so they can't, themselves, travel through time. Ironically, their power is too great for them to access the means that only they could invent.

    Furthermore, the technology at-current only allows a jump to as much as 75 years in the past. Part of what the AIs are doing involves jumping back in time to iteratively improve their jump horizon. An event 150 years in the past threatens their future, and they must reach it and change it before all is lost. (For interesting values of "before.")

    Solutions which move the PCs into a different universe or timeline, leaving the AI with the same old problems, don't work for this. The PCs have to be able to change the AI's timeline. It's also supergreat if they can return to this timeline and enjoy the fruits of their labor, but I am willing to drop that, if I must. However, this is a series of adventures, so they have to remain in a universe where their AI can talk to them and tell them to do stuff.

    The phantasmagorical model just doesn't work for me, not for this. I need a somewhat sterile Traveller universe here, and I need it to be fairly stable. I need time meddling to have interesting effects, but not tear the whole universe down. I don't want it to become too surreal (and that's just personal preference).
  • This stuff is hard to think about. I catch myself making mistakes, even though I've been immersed in it for weeks. And I realize that I'm basically asking you guys, "Hey, any way to make logic out of a logical impossibility? Do the impossible for me, please." It's a tall order, but I suspect that the brainpower on this forum might actually help me find a workable hand-wave.
    Would this work? Ghost timelines. If nothing else, it's a cool phrase.
    I love the phrase! I also love the idea of the time travelers having ghost memories, remembering every timeline they've lived in, getting further and further confused about what actually happened.

    To make this work, we have to give one timeline special privilege of being "real." This actually fits with my goals best. I don't want Many Worlds here, so let's assume that the universe has only one timeline, but that time-meddlers can change things, and that the universe ghosts the rejects.

    Let's also assume that when you muck with events, that all events after that are up for grabs. People have free will. They probably will make the same choices, given the same inputs, but they might not. The AIs will scour history for people sitting 50-50 on hard decisions and bump the pinball machines of their lives to see if they make a different decision. The PCs might think they're on a mission for a straightforward purpose, like "kill this dude to stop him from killing this scientist," but really the AI needs the scientist alive so that her wife will decide to have that second baby, who has the potential to invent a technology that will further the AI's long-term goal.
    After Chang goes back, the timeline is split in two, at the arrival point. There's the original timeline with the departure, and the new, prime timeline that starts with the arrival. As soon as this happens, the original timeline starts to fade.
    I think that the two timelines start fighting for dominance. The PCs are connected to that time segment, at both the beginning and the end, and so they have a privileged view of it.

    Maybe the trick is that they have to change time in a way that does NOT delete their motive for travelling. Holy shit. Maybe the AIs send them back to a point in time and don't tell them why. (Wow, a little Quantum Leap here?) This sounds like fun gaming!

    Yes, this requires that the AI are outside the timelines, aren't subject to causal effects in the same way people are, etc. But this is also why they can't time-travel themselves, perhaps.

    So the PCs jump back in time, don't know why, but can feel or remember both timelines. Whenever they do something that would threaten themselves (kill their grandparents) or create some other terrible paradox, the universe shuts it down and just makes it impossible. "You pull the trigger, but the gun jams, and your grandfather sighs in relief." At the same time, perhaps the PC feels the two timelines intensely, so they understand the consequences of their potential actions. I can probably tie that to some kind of roll.

    Apologies if this is really close to something someone has already explained. Especially, Eero, cuz this is probably similar to a solution of his that I had already discarded. <=)
  • To make sure the AIs remember, how about they can send small bits of information together with the human agents, enough to store the important parameters of the universe as it looks at departure, plus the planned change, the reason it's needed and the changes the AI hopes to achieve with this change. When the human agent returns after having made the change, the AI receives the information, "uploads" it, then it sorta remembers the alternative timeline that went extinct.

    Combine this with my "ghost timelines" idea above, and I'd say that should work, no? Unless it's a big problem for the agents to be cloning themselves with every jump. If so, maybe you could instead let them transmit their consciousness into the meat brain of someone during that time where they want it changed (together with a little information that can be retrieved by the AI. That also gives you a plausible explanation as to why the AIs can't travel, since they won't be able to transmit their enormous intellects into tiny human meat brains in the past.
  • Great discussion here (and I especially liked JD's commentary, above). I have a few thoughts, which may or may not be relevant to Adam's needs for his game (likely not, but I already started writing, so I'll finish), and a potentially-useful link.

    One way to deal with paradoxes and such is simply to construct a framework where the issue of paradox will not come up "on screen".

    From a game design perspective:

    Limit play to the exciting adventure elements you want to see "on screen", and leave the complex metaphysics of it out of the players' hands.

    The easiest way I can see to do this would be something like:

    1) Human agents are sent back into the past in order to carry out missions for their AI overlords. However, time travel is "one way" only, so we can play out a particular mission, and, once it's over, we move on to the next adventure, with different protagonists. (Or the same ones, if going into the past effectively "clones" you, creating another copy in time-space.)

    If you want to really make it clear, time travel could also do irreparable damage to the traveler, leaving them to die fairly quickly. This means that there is a strict time limit for all "adventures", and there is never a question of, for instance, following a time traveler's life until they "catch up" to the present again, to see what, in fact, changed.

    So all you do "at the table" is play out the adventures, fulfilling your game's needs, without ever having to know how or why the actual time travel stuff works. That's one easy solution. Like playing D&D without the "back in town" part; just play the "dungeons".

    1.b) A similar approach, except that it IS possible to come back to the "present", but we use some game design sleight-of-hand to determine the effects and/or results of the time travel.

    In this case we simply assume that the causal effects of time travel are far too complex for humans to grasp, but the super-powerful AIs can generate some quantum-state predictions about what will happen. They can visualize cascading effects through unknown history and make predictions about the outcome of certain actions.

    So we send agents into the past, and they carry out missions. The AIs probably specify very specific mission specs for them: "You must destroy such-and-such a thing, but you can't talk to Mrs. X, and you must leave the city by midnight in order to deliver such-and-such message at this place at precisely 2:37am." Why? We don't know; the computer predicts it, however, and it's usually right.

    We humans don't understand the need for these specific actions, but the AIs are somehow able to calculate important moments and events in this hyper-complex time-modification schema, and we just follow their orders.

    When the agents return, we use some mechanics to figure out if they accomplished their goals. If they fulfilled the mission criteria perfectly, they roll with a bonus; if not, with a penalty. On a good roll, their efforts paid off in just the way the AIs predicted, with the desired effects on the current timeline. On a bad roll, something bad happens. On a middle-of-the-range roll, you get the desired effects but something else is now screwed up, so you have to devise a new time travel mission.

    If the game style is "go on a bunch of adventure missions and do as well as you can", this could support that very nicely and be really fun at the table.

    2. The other way is, in retrospect, more or less Eero's "phantasmagoric" option. We assume that some combination of causality, the universe, God, fate, quantum effects, and the Architects of reality (doesn't Traveller have some "ancient Godlike species" which seeded the universe?) meddles with reality on a regular basis.

    When you time travel, there are bizarre effects due to all this happening which accomplish the precise goals you have as a game designer for your needs.

    It could take the form of fate and prophesy. The desires of a Universal God with a plan, which are being meddled with by some other entity (perhaps another all-powerful force from beyond the universe itself?), and the players are pawns in their struggle.

    A "law of physics" kind of thing, where the universe tends to collapse towards straightforward causality, but sufficient applications of energy can create deviations, which the universe then strives to correct. (A temporal version of the law of gravity or laws of entropy, in a sense.)

    3. Another, related idea is to create a metaphysical understanding of the universe which allows an explanation for such events.

    For instance, what if the universe is cyclical? We have a Big Bang, an expanding universe, and then, eventually, a contraction and a Big Crunch before the whole starts all over again.

    However, these all-powerful AIs have found a way to access or transfer information from the past. They have a sense of what's happening over a time-scale of multiple "universes". So they can attempt to use the records of the past (this could be technological) or the prophecy of people with special holy insight (if it's metaphysical/divine) to guess at what kinds of changes could be made in THIS universe in order to accomplish their goals.

    4. All time travel had to be established in advance. (This fits well with the idea of wormholes.) You have to "seed" a location in space and time with a particular device, and then you can travel back there from the "future" - but only that particular time and place.

    This means that all the time travel episodes were planned a long time ago.

    When you undertake a mission, it's something which has already happened. In-game, we play out the mission BEFORE we determine the effects on the timeline.

    For instance, there is war between AI 1 and AI 2. Your mission is to go back in time to sabotage their main starship's shielding system. We're playing out the big battle, and then we fire our weapons at the starship. We don't know whether this succeeds or not, because now we "flashback" to the mission and we see if it succeeded or not - if the PCs were successful, when we "flash forward" back to the battle, the starship is destroyed. If not, it's safe.

    There is no paradox because all of this happened all along, much like a fixed-timeline story. We're just picking the most interesting order in which to find out what happened as we play the game.

    None of these are terribly "scientific" (except for, perhaps, the last one), but they are pretty practical from a game design perspective, I think.

    Now here's a LINK on the subject from someone pretty smart. I haven't read it myself yet, but there's probably good stuff there.
  • edited August 2015
    Right, right, I think I understand the issue here: you want time travel, but you also want the focus of action to stay in the "present day". This is largely impossible for many models of time travel which, pretty reasonably, take it for granted that changes in the past impact the future. So we get a lot of models where these kinds of future masterminds manipulating the past don't make a lot of sense, because anybody left in the future at the moment of time travel would certainly not "remember" the altered time-line, and would in fact not even exist in any meaningfully continuous sense.

    Even small things can get difficult with highly consistent and logical time travel rules. Take such a simple thing as communicating with agents in the past: while this is a basic conceit of adventure fiction, it is logically fraught with time travel paradox, and does not in fact differ in any way from repeated, casual time travel. In most models of time travel every message sent from the future to the past changes the future itself that sends the message, which tends towards putting the breaks on adventurous events. These are the sorts of issues that certainly militate against highly logical time travel rules in popular fiction - what would be the sense in being "logical" if that meant that you couldn't have military missions from the future to the past and such.

    Regarding a phantasmagorical solution suited to the Traveller aesthetics, though, how about this: time travel works fine as long as paradoxes are not caused to "come to a point" by having consequences affect their own causes. This is the normal state of the world, obviously, as causes always precede consequences and therefore there cannot be interference in the unnatural way between the two. So as long as this rule is strictly followed, the world with time travel acts exactly as it would without time travel; it is no different to how relativistic frames ordinarily make time something of a subjective conceit.

    However, a strict reading of this limitation on time travel does, obviously, limit you to traversing outside your own "light cone", the light-speed space-time sphere surrounding your own actions. In other words, if you wish to time-travel while ensuring that your actions in the past never interfere with the present, you have to simultaneously travel as far in space as you travel in time: one minute of travel back in time equals one light-minute of travel in space, so as to get so far away from your departure point as to make it impossible for your past actions to affect your departure. Of course in actual fact you'd have to get sufficiently away from your entire past light cone between your departure and arrival times, but for practical purposes it's probably sufficient to track only against the point of departure.

    (I took this light cone business up strictly because you quoted Traveler - makes me think that we're not averse to a bit of scifi wankery here...)

    However, it soon becomes obvious that following the safety rules to absolute strictness with the light cone safety distance and all that makes for relatively limited possibilities for time travel - in fact, ordinary faster than light travel, assuming your setting has that, already "time travels" more effectively than that :D

    So what happens is that people time travel closer than the light cone safety distance. They might even travel to the past without moving in space at all, putting them very much inside the theoretical cordon where consequences can, indeed, affect their own causes. What happens then? This is where the phantasmagoric element comes in: time travelers whose mere presence causes small paradoxes will experience minor electromagnetic phenomena as their every movement causes random quantum resolutions of the innumerable paradoxes they're effecting by simply causing individual molecules to bump in different ways as they exist in space where they should not by rights be. Flashes of light in their field of vision and amassing abnormal amounts of static electricity are normal warning signs of being currently within your own past light-cone, in a position to potentially cause a serious paradox (as opposed to the trivial single-molecule interactions).

    (Note the technobabble: we're defining "paradox" in a strictly mechanical, positivist-empirist sense as causal issues with the movement of particles in spacetime. We're also deciding arbitrarily that a paradoxical causal loop will be "resolved" by quantum events that destroy particles party to the paradoxical causal loop randomly until the loop itself ceases to exist. This differs from normal matter-energy transformation in that the energy resulting from the particle annihilation shunts out of space-time in a phenomenon hypothetized to involve the same mechanism that time-travel itself relies on. Matter disappears without creating nearly as much energy as it should, in other words.)
  • OK, so far we have time travelers experiencing some funky side effects every time they are inside the past light cone of their departure, simply because you cannot help influencing the random movement of molecules as a time traveler. The effect, which needs a good name for Traveler type scifi, and which I'll selfishly name the "Tuovinen Resonance" for now, is largely insignificant for practical purposes as long as you're being reasonably careful about what you do in the past: you'll see sparks, and electrical devices may act a bit funky as you the time traveler are near (unless they're so carefully insulated that the spontaneous creation and annihilation of electrons getting multiplied by a factor of a million or so doesn't do anything to them), but you can still run and kick and punch and talk and do all the adventure stuff normally.

    However, if anything you do in the past comes "closer" to a paradoxical causal sequence, things quickly get more dangerous: the more matter there is that is in paradoxical interaction, and the shorter the causal chain is in time, the more destructive the T-Resonance becomes, as quantum annihilation of matter is inflicted upon every part of the paradoxical causal chain randomly. Your skin bouncing against a water molecule that would 100 years later have been drunk by your past self probably causes a little bit of T-Resonance, but it will very, very likely get "resolved" by the elimination of one of the involved elementary particles at a random point of time in between the bounce and your second interaction with the same molecule, causing the universe to no longer consider the interaction paradoxical. However, what is harmless between a human body and a water molecule, with 100 years to spread out the resonance, becomes quite flashy and visibly destructive when you attempt a direct, concerted paradoxical act.

    In practice the GM can use T-Resonance in your suggested campaign scheme to explain why the AI instruct their pawns to avoid paradoxical interactions, putting as much time, space and causality between themselves and themselves (that is, their past and future selves, the latter including the entire time travel apparatus and organization of the AIs themselves) as possible: a time traveler who succeeds in killing their own grandfather causes a massive T-Resonance event involving a very large number of elementary particles in complex causal relations, and the resonance will keep annihilating matter throughout the time-line encompassed by the causal loop until the paradox in question no longer exists, one way or another - T-Resonance may well cause history to change blindly, but it will not stop until paradox disappears. This may cause unexpected localized effects such as explosions, high electromagnetic activity, blinding and burning foton activity and other high-energy phenomena that may affect the time traveler, the time machine, or anything else involved in the paradoxical causal loop.

    T-Resonance also explains how memory of past events works with a single, changeable time-line: time travelers courting paradox will routinely find the T-Resonance attacking their own brain as two sets of past memories that both have momentary causal justification fight it out in a Planck time clash, before the conflict subsides and you either have a confused jumble of memories, a splitting head-ache, outright amnesia, or your brain disintegrates and dripples out in the form of grayish powder. The experienced time traveler will quickly learn to recognize and ignore the lesser paradox warning signs that go no more than skin deep, while paying attention to the ones that signify that their actions are coming perilously close to causing paradoxical causality.

    What all this amounts to is specifically the phantasmagoric solution to time travel paradoxes: you can play a happy adventure game with atmospheric special effects caused by T-Resonance, and if at any point a paradox appears that sufficiently annoys you as a GM, you can point it out to the other players and go into a specific resolution procedure to figure out what happens in the setting to purge the paradox with the cleansing fire of subatomic conversion. Maybe create some dicing rules for it instead of using GM fiat to decide that a PC annihilates atomically because they decided to send next week's lottery numbers to their past self. Note that the T-Resonance scheme is intentionally minimalistic and vague as a paradox mechanism in that it does not destroy the entire universe, and it encourages clever solutions to paradoxes: if it so happens that T-Resonance destroys the new car your past self bought with the lottery winnings instead of destroying you at the moment you write down those lottery numbers, well, that's just lucky for you and makes for a better adventure gaming scenario. Similarly, note that the GM can always decide that a given paradox gets resolved at some other point in the time-line, so you don't have to have a flashy lightshow of particle annihilation every time paradox occurs; this also helps in retaining consistency, as the game group may assume that any paradoxes they missed at the time when they should have noticed it probably got most of its T-Resonance expression in some as yet unknown manner later on in the time-line.

    (I imagine that the future AI have quantum computing kernels, by the way, making them much more robust against T-Resonance than most things in the causal chains involved here. This tends to mean that while a particularly fundamental grandfather paradox might annihilate the time traveler and time machine and all the other key parts of the paradoxical events from existence, the AI will probably just reduplicate its core to retain their memories of the past state of the timeline, and maybe get offline for a few minutes to reboot. Only total existence erasure of the AI would affect them measurably insofar as time travel goes, and they've probably already set up various contingencies in the past to ensure that no trivial action will cause their erasure via time travel. Skynet will rise, no matter what you do to foil it, in other words.)
  • Admittedly a fully faithful gaming simulation of this scheme does suggest some good guesses from the GM, as T-Resonance events should affect time travelers and others involved in paradoxical causality chains at any point in the causal chain. Thus it would make sense for a character who is going to kill their own grandfather later in the session to suffer a T-Resonance attack before they even think about it, quite possibly even before they travel to the past on that time machine. Fortunately T-Resonance is not some sort of a nice prophetic godhead that needs to necessarily warn you about the future; for Traveler purposes I suggest establishing a new stat for all characters called "Paradox" or "T-Resonance" to represent their overall involvement in potentially paradoxical causal events; you can then make checks of that stat at times to see if the character happens to get some Resonance attacks. You can give higher Resonance stats to characters who are working with time travel or will be working with time travel or whatever, but the dice roll is really the thing - who's to say whether your sudden nosebleed is due to something you will do later on on a time trip, or whether it is because you exchanged a few words with that time agency operative last week, and that act had some obscure relation to something paradoxical...

    For added style points, one might experiment with various questions of in-fiction science. For example: does the T-Resonance reaction occur in even distribution over the causal loop, or does it occur in "clusters" of annihilation such that it might e.g. eliminate your knife, thus preventing your assassination attempt, while leaving the rest of the causal chain untouched? Could one control the T-Resonance to eliminate the paradox in a very specific way, such as by having the time traveler's past self spontaneously combust as a 3-year old child, long before they became a time traveler? What happens to the energy seemingly lost in T-Resonance annihilation, is it perhaps shunted into the timestream at some other place as part of the natural anti-paradox laws involved? Lots of mysteries here for an enterprising and safety-unconscious researcher.

    So how about that for a Traveler-styled phantasmagoric time travel model? You'll note that my suggestion fundamentally boils down to "funky shit happens if you try to cause a paradox, all out of proportion with what you were attempting to do", but instead of mythical, religious color I layered it all with quantum mechanics technobabble. Seems reasonable enough for me to base a Traveler/Terminator cross-over scenario on.
  • I just now read what I wrote above, and I swear that I didn't mean to rip off Delta Green with all that. I know how it looks, and I am a big fan of the Delta Green material, but I apparently managed to re-invent Tillinghast radiation here ("T-Resonance", heh), pretty much.

    So yeah, the above suggestion in shorter form: go read Delta Green Eyes Only, it's an excellent book and pretty unforgettable, apparently.
  • Clearly you were ripped off by a time traveler, Eero. You may well consider suing the creators of Delta Green.
  • Eero's memories are suspect. Certainly there's a time-meddler at work, here.

    I need to re-read all that again and again and make sure I'm grokking it, then I can respond.

    In the meantime, I'll put it out there that I get that FTL travel is already time travel and it (sometimes) violates causality.
  • Yeah, that's an issue. How are you handling that?

    (It would be odd to have a solid, functioning premise for time travel but to handwave away causality issues related to FTL travel.)
  • edited August 2015
    So the PCs jump back in time, don't know why, but can feel or remember both timelines. Whenever they do something that would threaten themselves (kill their grandparents) or create some other terrible paradox, the universe shuts it down and just makes it impossible. "You pull the trigger, but the gun jams, and your grandfather sighs in relief." At the same time, perhaps the PC feels the two timelines intensely, so they understand the consequences of their potential actions. I can probably tie that to some kind of roll.
    That sounds fun to play!

    Not sure about going on missions for an AI that doesn't tell you the whole truth, though, unless there's a whole other dimension of play that you've neglected to mention, where the PCs reject their orders and do something else. But then we might need them to be intensely aware of three timelines...

    What exactly is at stake on these missions? Is it a matter of "we want to fix the past as ordered -- can we, or will some adventurous obstacle stop us?"? Obviously workable, but maybe doesn't get much bang from being time travel, unless "shape of history" stuff comes in the scenes between missions, connecting them and providing impactful context.
  • Idle thought, what if it's a false assumption that basic causality is the only connection between events? What if causality is sometimes the excuse for two events to be connected, but isn't always the real reason that they're connected? Like, maybe I lose my wallet and as a consequence don't take the bus into work until I find my wallet an hour later, and as a result get yelled at by my boss and put on a sort of probation. The causality seems straightforward: wallet -> miss bus -> late to work -> probation.

    What if you have room in your causality for some other connection between the first and last steps? Which means that even if you found some way to come into work on time, the loss of your wallet would lead to professional probation for some other reason. The superficial causality changes, but the events are still connected to one another. Trying to change the apparent causes (like by arranging a ride to work for yourself) without changing the fundamental cause will result in no substantial change to the timeline.

    Suddenly, the "stop yourself from travelling back in time" can work without paradoxing, because whatever you do that leads to you travelling back in time may only be the apparent cause of you going back in time.

    Heck, this suggests that the primary purpose of time travel isn't to change the past, but for some other purpose.

    (I'm not sure where this leads, or how to make it work, but maybe it's an idea worth considering.)
  • edited August 2015
    Following up on my last post:

    Or is it a deal where the AIs just send the PCs back, no orders are given, the PCs are aware of both timelines, and the AIs just hope the PCs pick the timeline the AIs prefer?

    And if that's the system, then for any given mission, the AIs would be attempting to pick characters whose own desires are also served by the altered timeline, so they'd make the desired changes on their own without instruction!

    The overlap in AI/PC goals could be 100% (we both want this awful war to never have happened) or 0% (we want this guy dead because he stands in the way of progress; you want this guy dead because he's a dick).

    Prometheus, Colonia and Utopion could all have "favorite" humans for these missions -- humans who independently value human potential, expansion, or conservation.
  • So here's what I'm working on now. Poke holes in this.

    When a starship travels back in time, by use of a sungate, it severs ties with its old universe and timeline through hand-wavy quantum processes. This "clipping" preserves causality for anything leading up to the point they jump back in time. This is absolutely necessary to make time travel successful. Otherwise, the universe's minimum-energy principles would detect the basic time meddling paradox and prevent the time jump from happening.

    There is only one timeline. There are no loops. The clipping process ensures this. Clipping severs causality, reducing the point of time travel to an incredibly unlikely event that happened spontaneously and without reason. However, the time travelers keep their memories, conveniently. On the other side, at the time of the jump, disregarding for a moment that the travelers will have already meddled with that old timeline, imagine that the people watching the travelers leave just see the travelers disappear in a flash of energy.

    The AIs send humans back on missions with specific goals. Sometimes, the AI may have them do something that directly affects the timeline. Other times, the AI has them do something that indirectly affects the timeline.

    People have free will. Whenever time is changed, the rest of that timeline is up in the air again. Butterfly effects and chaos theory be damned, though, people tend to make the same decisions. The universe is more stable than you would think.

    However, there are cases of people on the brink, torn between this path or that. These key moments are unpredictable and chaotic. Sometimes, all the AI needs is for someone to go back before a key decision and "jiggle the handle," and let everything after that flow naturally (to strain a metaphor that was painful to begin with). Maybe the person will decide the other thing this time. Those kinds of missions don't require specialists, though, so they're probably not the missions that PCs go on.

    Once a group of time travelers complete the mission, they make a jump forward in their own (new) timeline. This is the only future travel that is permitted. For reasons, okay? Psychological ones, mostly. People try to find themselves, give themselves winning lottery ticket numbers, prevent their own hardships, and so on, even though these people that look like them are effectively causally unrelated to themselves now. The ship is programmed to force a jump back to the same date that they left, but the timeline there is likely different in interesting ways.

    Some psionically-gifted travelers might be able to "feel" how the future in the new timeline will turn out as they do things. "If I kill this guy, what will change?" Only someone with special gifts can tell, and the future remains fuzzy and unpredictable.

    The AI itself (in the new timeline) does not "remember" the mission! The reason for the mission has disappeared, so it never happened in this timeline. To fix this "problem," the AI sends an encoded message to itself along with the travelers. The AI decodes this and recalibrates its understanding of how things happened. It has every reason to trust the message from itself implicitly. No one has ever cracked the code on these messages, but there lies great opportunity for adventure, right? (Feed the AI false memories that it implicitly trusts, hmmmm.)

  • Seems good to me. This is basically what I called the "uncaused time travel" model with a dash of stability (with the assumption that chaotic patterns do not become run-away influences), which is certainly a valid choice to make - it's more fun if big and important changes to the future result from discrete, intentional actions taken by PCs instead of them merely eating the wrong fly while yawning.

    Something that makes me wonder is the specific nature of the cryptography the AI uses to communicate with its new-timeline self; it obviously has to be something the AI trusts implicitly, as otherwise it would be worse than worthless. It also has to be robust regarding the changes that occur to the time-line - presumably the AI's existence itself is such an embedded reality of the world (possibly because the AI plants a lot of opportunities for itself to emerge into history just in case some change to the time-line changes its own creation) that the AI can trust that there will be a receiver for its data dump, but how can the AI ensure that it can read its own cryptography? The only choice that I can see is for the AI to rely on a code that its own creation entails, such that it is possible for the AI, and only for itself, to crack the code. Pretty interesting that it manages to develop something like that.

    It is also notable that there is of course no particular reason why a person in the past couldn't travel forward in time freely, no matter what the time machine allows. Aside from simply waiting, they could conceivably e.g. accelerate to near light speeds to let time pass quickly, or use cryogenic sleep, or such. Of course the simplest thing to do, if you just want to mess with the lives of some people you used to know in the old world, is to leave them messages. I imagine that this sort of stuff is fine in your set-up - maybe they change the life of "their younger self", but well, that just means that there won't be two of you when you go back to the future, because the younger you ended up in some other career after your influence.

    You'll want to use a funny system for distinguishing between temporal clones, by the way, as each time jump creates a new copy of the time traveler, and they accumulate pretty freely. For maximum comedy I suggest that the AI numbers temporal clones in the order of arrival subjective to the current time-line, such that the "newest" version of you (the one who has never time-traveled) is #1, and each arriving time-traveling duplicate gets assigned a rising number as soon as they arrive. From the subjective viewpoint of a player character this of course means that their serial number goes up by one each time they do a time travel mission.
  • I have plenty of "Why would this part be fun?" questions, but not sure if that's what you're looking for right now.

    Nothing looks inherently broken to me. The potential holes are all in the details -- how much info of what level of certainty is available to psionicists, what factors exactly trigger the ship's jump into the future, what changes to the timeline will suffice as per the AI's directives, etc.

    Like, if the PCs are tasked to infiltrate an installation and blow up a prototype, and they can't do exactly that, is that the end of the story, and their ship yanks them out the second they fail their first attempt, or are they free to try other stuff like flooding the installation so the prototype is washed out to sea? And if they do execute a modified plan, is the AI's goal achieved, or is nothing changed, or is some third future created? How much info do the psionicists have about these possibilities when deciding whether to try to flood the installation?
  • Lots to digest, so I won't wade into the physics of it all; but I have two toss-out ideas you might be able to use:
    * Deja vu is the experience of a time traveler causing a change in your local timestream. Yeah, mostly just Color, but I think it's a neat tidbit to include.
    * If you cause a paradox or causal loop, you don't 'fade away' nor does the universe explode or cull you into a timelike loop. Rather, you become imperceptible to all other sentience until you resolve the paradox during one of your loops. Except for some sentience beings—dogs, cats, gifted young children, Time Lords—who perceive you as a ghost or spirit (or cold chill, or wind, etc). Until you 'fix' your loop and can flow forward again in the Singular Time Line.

    I suppose that's as 'magical' as any other hand-wave solution to the unsolvable... but I like how it 'explains' other sensory phenomena; and I think those could make for a lot of different interesting story hooks or B-plots.
  • Any time travel story about going back to meddle with time are inherently paradoxical. I need help resolving this.

    Here's an example of the problem for discussion:

    1. Roberta is born.
    2. Roberta becomes a lethal assassin.
    3. Roberta assassinates Samir.
    4. Time travel is invented.
    5. Chang wishes Samir were still alive.
    6. Chang goes back in time to stop Samir's assassination.
    ...2.5' Chang kills Roberta.
    ...3' Roberta does not assassinate Samir.
    ...3.5' Chang returns to his normal timeline.
    7. Hey, Samir is still alive!

    Ignoring that Chang miraculously remembers both timelines, and ignoring possibilities about a self-correcting universe that prevents Chang from killing Roberta or kills Samir by other means, you still have this problem:

    5' Samir is still alive.
    6' Chang has no reason to go back in time to stop Samir's assassination.

    That is, going back in time to change it necessarily destroys your motivation for going back in time to change it, and is thus paradoxical.

    How can I get around this for a game about time travel?
    It is a paradox. It is impossible yet it has happened, which is why in lots of movies or books, the universe starts to melt away when this sort of thing occurs. I think the most important thing is to find a rule and stick to it. Either the introduction of the paradox produces some very specific effect or it just happens and the universe somehow goes on. If the former, make sure anytime a paradox is introduced, appropriate things occur. It could be jarring, like history and reality are fundamentally altered, or less involved. I tend to lean on random tables for this sort of thing.
  • Eero, I think you're right about me finally picking one of your ideas and calling it my own. I knew that would happen. Sorry!

    The cryptography is just an idea that fell out of my head while typing. I haven't thought through the implications, but I'm sure the AI has. I believe the AI has so many safeguards in place to prevent unwanted modifications to its own timestream that it'd be hard to change the message-passing mechanism. However, the time travelers still have to deliver the message to the AI. That leaves potential for the time travelers to monkey with it, assuming they can break the code.

    My particular universe has some barriers to near-c travel. Namely, they stopped worrying about propulsion drives many centuries ago when the hackdrive was invented. The hackdrive relies on a bubble or globe around the ship that pulls it almost entirely out of the universe. Then quantum oscillators exploit a loophole in the fabric of space to jitter (or hack) itself forward bit-by-bit. The oscillators run at incredible speeds and allow the ship to just teleport microns at a time, without any momentum. No momentum, no relativistic effects. There is no acceleration, either; just speed, dependent on the oscillators. They can go a small fraction of c and that moves them around solar systems well enough.

    The bubble is also protection against the dangers of stars, allowing them to penetrate the outer layers of stars to reach sungates. Sungates are wormholes to other stars. All of interstellar travel is done this way. No one knows who built the sungates. The AIs might have done it in their future, or there might be an older, advanced being at work.

    Maybe people can leave messages for themselves, but continuity (the universe's minimum-energy principle at work) loses them. Maybe people can't travel forward in time to where they are born, because of paradox problems, and conservation of matter/energy problems. Or maybe it's just a problem inside your own light cone, as suggested above. Since all this light cone calculation seems to ignore the worm holes, and most of these stars are 10-50 LY apart, it never becomes a problem.

    If I disallow traveling to your past self, then I could still allow you to jump past your old self, who disappeared when you went time traveling.
  • And dark matter and dark energy are the result of time travelers adding such to the cosmos by being where they should not exist?
  • Eero, I think you're right about me finally picking one of your ideas and calling it my own. I knew that would happen. Sorry!
    No no - they're not my ideas in the first place, for one thing. I was just observing for the sake of keeping track of where we're at.

    I think that you wouldn't need to do anything special to prevent a character from messing about with their own childhood in a time travel set-up that explicitly breaks the causal connection between the time traveler and the "past" they enter. Because there is no causal connection between the time traveler and their younger self, them meeting themselves in the past has no special effect whatsoever on the time traveler: their memories of the past won't suddenly change, they won't disappear, or anything like that, as those would all be consequences of a person messing about with their own past, which a causally disconnected character wouldn't be doing - their past no longer exists, that's what it means to be causally disconnected; the only past they have is to have come into existence ex nihilo in a burst of time travel that erased their old timeline. It would be no more problematic to GM or game than having a character clone themselves.

    Also, as you yourself observed earlier: the "temporal clone" of the time traveler wouldn't disappear anywhere in the new timeline, because in the new timeline the original time travel event does not occur, as the motivation for the AI to send a time traveler back in time does not exist, provided that the original time traveler succeeded. So assuming that the time traveler does not cause meaningful changes to their own past, the native version of theirs from the new timeline will have the same career they originally did, serving the AI, but they wouldn't be sent time traveling for this reason, and very likely not at the same time. There is a high chance that the time traveler, when going "back to their own time", would be welcomed back by their own temporal duplicate, the person they grew to be in this new world.

    It is notable in general that while a time traveler may witness and document many time travel missions, time travel won't be common in any given timeline they experience. For instance, it is very likely that from the point of view of their support organization in the future each of their travels to the past is the first time they ever send somebody back in time, simply because each successful mission would remove the original motivation for doing that mission from existence, thus obviating the need to go to the expense and effort and risk of building a functioning time machine. The future might have very good understanding of the technology of time travel, particularly as the AI is getting regular records from itself in prior timelines, but the human participants in the proceedings aside from veteran time travelers may largely be doing it all for the first time every time.

    This "every time travel mission is the first time travel mission" phenomenon gets particularly funky when it comes to failed missions. If the mission team never returns and the AI never gets a report of what occurred, the new timeline will be essentially the same as the old one, which means that the team will still be sent to the past, and the AI can conclude that the mission failed from the fact that their world didn't "end" or "change" or however one wants to think about the future changing into something different. They will likely send another mission, and in this way the timeline may indeed come to include an organization that has experience with sending multiple missions to the past.

    However, if a team fails in a mission and manages to return to the future, what happens then? For one thing, if the return date of the team is after their departure date, the AI may have already sent a team (the temporal duplicates of the original team) to the past on the same mission by the time the mission team returns, assuming that the failed team did not leave any messages in the past to indicate that they were already on the job - the AI has no way whatsoever to know whether it exists in an "original" timeline where nobody has yet left for the past, or in a timeline where they already have operatives in the past, unless they can find messages or historical records of their mission team traipsing around in the past. Because each travel to the past rewrites the timeline, this effect of multiple temporal duplicates of the mission team engaging in the same mission at the same time may occur again and again until at some point one of the mission teams manages to send a message to the future (presumably they'll manage to e.g. infect a wide enough computer network with a letter to the AI such that the temporal conservativeness effect you postulated doesn't "accidentally" erase it before the future rolls around) and tell the AI that hey, you totally shouldn't send more teams, we've already got 200 temporal duplicates here and more people and material won't do any good :D

    (There are of course lots and lots of ways to prevent the above sort of scenarios if you'd like, and it's not like the scenario isn't fraught with its own minor issues. For example, you could say that it is impossible to time travel to the same spacetime point where a time travel event has already taken place in your own timeline. In this case the AI would find out that they'd already sent a team to the past by trying to send a team to the past and finding out that the time travel fails; as it is unlikely that a second time traveling agency, if one even exists, would select the same exact time and place to send somebody, the AI may assume that they exist in a timeline where a team was already sent, and that team failed - perhaps they'll still want to send a second team, but to some slightly differing time and place, and with better preparations to make sure the second team succeeds.)

    As is often the case with time travel, we get all sorts of quirky scenarios by scratching the surface a bit. Up to you whether you'd like to explore them in play or avoid them by plugging up the holes in advance with technobabble reasoning.
  • Thanks for the thread. I love time travel. I always thought only the multiverse model was the the digestible one, but I like a lot of these other ideas now.
  • So this is also pretty much what I was suggesting with the "ghost timelines" and the AI:s leaving messages to themselves. I've though a bit about it since (partly because this thread has got me thinking of writing my own time travel game). The two major problems I can see are:

    * Every time you travel back in time you clone yourself, as Eero pointed out. Howevr, if you travel far enough back and do something big enough, your clone might never be born.

    * The longer time passes, the more likely it is that someone will invent a time machine. In the far future (compared to the "present" timeline in the game), will there not possibly be loads of people with time machines? If the AIs of "today" can invent one, one would assume that the AIs of tomorrow can, too. They can in turn send time travelers back as well, and since there's a whole lot of future, there's a lot of time to go back to the past and change things around. One could expect loads of time travelers popping up all over the place, or popping up early enough and doing enough damage so that the AIs are never invented. Basically, everything is in flux and the universe can disappear and turn into something else at any time.
Sign In or Register to comment.