How might the occupants of any given universe identify which (if any) of the many proposed Time Travel Paradigms govern their potential use of time travel? By "Time Travel Paradigms" I refer to the proposed solutions to paradoxes attributed to traveling backwards in time, such as:

  1. Anything done while in the past will be revealed to have already occured in the past, before the time travel was attempted.
  2. The moment you arrive in the past, the timelines fork and you now exist in an alternate timeline.
  3. You simply can't.

My question is less about how the person travelling back in time would recognize the rules of time travel in their universe, but how that person (presumabley a scientist of sorts) would be able to decidedly prove that the universe works in that particular, definable way to the general/scientific public.

It would seem that the bare minimum would be the ability to travel back in time again and again, while producing the same results each time. Then, of course, multiple sources would need to independently verify the accuracy of these results by going back in time themselves. All in all, it would add up to quite a bit of time traveling with uncertain consequences.

Additionally, it seems that certain methods would be difficult to discern from others. For example, considering the two examples I listed above, assume I traveled back in time in an attempt to kill Hitler. I fail, and return to the present. Nothing is noticeably different, but I have no real proof whether my assassination attempt happened in the previous history of my universe (the first supposition above), or if my act split the timelines and I now exist in a very, very similar alternate timeline (second supposition).

  • 6
    $\begingroup$ Because we are talking about a potentially very exacting topic, would you mind if I quibbled on wording? Science has never once "decidedly proven" anything. It provides high statistical certainties, but never 100% (this is also an essential trait required to allow science to adapt to new discoveries) $\endgroup$
    – Cort Ammon
    Jul 10, 2015 at 22:20
  • $\begingroup$ @CortAmmon A bit captious though isn't it? Can you even use the word "never" with that mindset? It seems every absolute statement must be qualified with "probably" or "most likely" in that case. Science has made a lot of valid axiomatic statements, fairly so. Is it not a given that all human statements are, at best, little 't' truth and never capital 'T' Truth? $\endgroup$
    – Samuel
    Jul 10, 2015 at 23:49
  • $\begingroup$ This article on time travel models may be useful for seeing some of the possibilities that are out there. $\endgroup$ Jul 10, 2015 at 23:53
  • 1
    $\begingroup$ @Samuel I use never because science is so entwined in statistics that any serious discussion of provability (especially on the fringes, such as dealing with time travel) should be careful with the wording. I do however, admit that I'm pushing an agenda. There are those who argue science to be a wider philosophy regarding "falsifiability," but I've never found one that could explain themselves satisfactorily when pushed, so I hold to the definition of the scientific method I was taught in school, with statistics. I may be trolling to find such a person to challenge that position ;-) $\endgroup$
    – Cort Ammon
    Jul 11, 2015 at 0:16
  • 1
    $\begingroup$ Reading your answer, Samuel, leads me to a question about this question: are we seeking "definitive proof" for the individual time traveler, or evidence which can be given to a non-time traveler to prove the nature of the world? There are a lot of things that are easier to prove to yourself than to prove to others. $\endgroup$
    – Cort Ammon
    Jul 11, 2015 at 0:19

3 Answers 3


Travel back in time five minutes and say hello to your past self.

If you remember meeting yourself before you left, you have paradigm one.

If you don't remember meeting yourself before you left, you have paradigm two.

If you walked out of the time machine without time travelling, you have paradigm three.

  • 2
    $\begingroup$ I would argue that being able to go back in time and speak to yourself would be an example of paradigm 2. If you attempted such a thing in paradigm 1, some series of events would prevent you from talking to your past self (otherwise this would result in a causality loop of needing to force your past self to also travel back in time to greet his past self). Now if you don't remember this happening before you travel back, but once you start talking to your past self you begin remembering that it had already happened, that seems more similar to whatever paradigm was in Looper. $\endgroup$
    – Etheur
    Jul 13, 2015 at 14:38
  • $\begingroup$ @Etheur So it seems the experiment will be a useful one for any of the proposed paradigms. $\endgroup$
    – Samuel
    Jul 13, 2015 at 15:32
  • 2
    $\begingroup$ If you remember meeting yourself before you left, you might have paradigm one or you might not. ("You" might instead be the "other you" from paradigm two.) This experiment cannot affirm paradigm one, it can only affirm the others. $\endgroup$
    – wberry
    Aug 16, 2016 at 21:36
  • $\begingroup$ So, let's say I make a simple check-list table with two columns: the first get a check-mark right before traveling to the past, and the second right after saying "hi" to myself. I'm about to start the experiment, and I take one last look at the table - won't I see all the check-marks already there? What does that tell me about the paradigm (maybe I'm at paradigm 1, maybe I'm at the last branch of 2)? $\endgroup$
    – G0BLiN
    Jul 22, 2018 at 19:55
  • $\begingroup$ @G0BLiN That would be paradigm one. $\endgroup$
    – Samuel
    Jul 22, 2018 at 19:58

If you have a time machine, you can easily distinguish versions 1 and 2 by actively trying to create a time paradox. In universe 1, you will inevitably fail to do so, even if it can only be avoided by the most strange coincidences. You may actually be able to win the lottery by forcing a paradox if you don't. In universe 2, if you go back yourself, you'll find that you can change the past, but if you send something (or someone) back into the past to change things but don't travel yourself, then you'll find that the past has not changed.

Universe 3 will be identified by you not being able to build a time machine. But of course, that cannot be an absolute proof, as you might just not have found the way to do it. However, the moment you've got a working time machine, universe 3 is disproved.

  • 3
    $\begingroup$ One thing about paradoxes and strange coincidences. If you are trying to make paradoxes, the universe has only one person to stop to stop all the paradoxes. You are more likely to die in a car crash than win the lottery. $\endgroup$ Jul 10, 2015 at 22:13
  • $\begingroup$ @PyRulez: Well, that's a good argument for using an automated mechanism to force your winning numbers to be drawn. If the paradox is resolved in an unexpected way, it's likely only the mechanism that gets killed. $\endgroup$
    – celtschk
    Jul 10, 2015 at 22:15
  • $\begingroup$ Make the machine fragile. It should hopefully be more fragile than you are. There are also issues with repeatability (if you plan to make a bunch of paradoxes, that universe will be that much more likely to stop you know.) $\endgroup$ Jul 10, 2015 at 22:19
  • 1
    $\begingroup$ I actually thought of a time travel story were a time traveller who thinks they are in #2, but when there is this hooded dude who keeps trying to stop them from changing the past. The hooded dude turns out to be you from the future, stopping you from making changes so the universe doesn't (the universe wouldn't need to use strange coincidences if you are stopping yourself.) $\endgroup$ Jul 10, 2015 at 22:21
  • 1
    $\begingroup$ @sumelic: If you have someone with such powers involved, there's nothing at all you can say for certain, not even if time travel actually works, and not even about things not related to time travel (for example, is quantum mechanics actually a scam by time travellers from the future messing with our experiments and memories? Maybe they just want to obfuscate the real laws of physics so we don't invent time travel too early!) $\endgroup$
    – celtschk
    Jul 11, 2015 at 8:30

Although Samuel's answer is succinct, it is not entirely accurate. You already realize it, but it is worth repeating that for some situations the three possibilities are indistinguishable. For instance, if you travel back in time and produce an alternate timeline, from the point of the original timeline the result is identical to the proposition that time travel is impossible - any attempt simply causes the traveler to disappear. Likewise, if changes are propagated forward, all sorts of unintended historical effects are produced. Larry Niven's Law of Time Travel suggests that for this situation the only "stable" history is one in which some combination of historical accident results in time travel "never" being invented. Finally, of course, version 2 is indistinguishable from version 1 in the new timeline.

With that said, establishing the existence of time travel to any desired degree of confidence is fairly easy. You simply tell an investigator to walk to the room next door and pick up the locked metal box which he will find there. Three days later, tell him to open it and observe that it contains that day's paper, which you bought that morning, then put in a box and travelled back 3 days to place in the room. By being careful with your transport arrangements you can prevent paradox or closed loops. Repeating the process repeatedly will establish your bona fides to any desired degree of reliability, although the amount of care required to avoid the investigator cheating will be considerable. The second time you do this, the investigator is going to be tempted to open the box early and use the prescience he gains. There are ways around this, such as encrypting the box's contents and supplying the key on the day the box is to be opened, but this can produce some very tricky details to get right to both sides' satisfaction, and is very similar to the problem of transmitting encrypted data without getting spoofed by an opponent.

Note - in looking over the preceding comments, I find that I am in agreement with Cort Ammon concerning "decidedly prove", although my phrasing about degrees of confidence was made without consideration of his objection. Science doesn't work by proving things, but repeated confirmation can be "close enough for all practical purposes". As in, take a gymnasium, and line up 100 teenage girls on one wall, and 100 teenage boys on the facing wall. Have each group walk half-way toward the center each time a bell rings. Theoretically, the two groups will never meet, but after a relatively small number of bells they will be close enough for all practical purposes.

  • $\begingroup$ About your "teenager experiment": Since teenagers are not mathematical points, but extended bodies, they will meet, even theoretically. Their centers will not meet, of course, but they couldn't anyway, thanks of the Pauli principle. ;-) $\endgroup$
    – celtschk
    Jul 11, 2015 at 8:37
  • $\begingroup$ Transmitting authenticated and encrypted data that is both hidden from 'man-in-the-middle' attacks and provable as coming from a specific person is a solved problem in cryptography. What is still difficult is key distribution. $\endgroup$
    – Green
    Jul 11, 2015 at 12:38
  • $\begingroup$ @Green - Absolutely. In this case the problem is getting the key to the investigator at a time which simultaneously guarantees that the message is identifiable as "future" information, that the key was not generated after the current information became available (and is therefore an honest key) AND that the investigator cannot use the decrypted information early enough to violate causality. As I said, "very tricky details". $\endgroup$ Jul 11, 2015 at 17:40
  • $\begingroup$ I saw this NIST site in someone else's answer. It's a really nice way to ensure that someone has knowledge of future events in a publicly verifiable way. nist.gov/itl/csd/ct/nist_beacon.cfm $\endgroup$
    – Green
    Jul 11, 2015 at 21:11

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .