Comment by cosmos0072

4 days ago

I have a degree in theoretical physics, and also did research on general relativity.

The result is cool, but it's not directly applicable to the traditional (sci-fi) scenario "I travel to the past and meet myself / my parents / my ancestors"

The reason is simple: the authors suppose a CLOSED timelike curve, i.e. something like a circle, where you travel back in time and BECOME your younger self - which by the way only exists because you traveled back in time in the first place.

A slightly different scenario would be much more interesting, but my guess is that it's much harder to analyze:

a NEARLY closed timelike curve, which arrives from the past, coils around itself one or more times - like a coil, indeed - allowing causal interaction between the different spires (i.e. one can interact with its future self/selves and with its past self/selves), and finally the last spire leaves toward the future.

> The reason is simple: the authors suppose a CLOSED timelike curve, i.e. something like a circle, where you travel back in time and BECOME your younger self

Exactly. This part of the paper is not really surprising or newsworthy. If you apply periodic boundary conditions, you get periodicity, duh. In the case of CTCs, this has been known for a long time[0].

> A slightly different scenario would be much more interesting, but my guess is that it's much harder to analyze: […]

Agreed. The only result I'm aware of in this context is a paper from the 90s by Echeverria, Klinkhammer, and Thorne about a thought experiment (Polchinski's Paradox) involving a billard ball entering a wormhole and colliding with its past self. Wikipedia[0] gives a good overview of the result.

[0]: https://en.m.wikipedia.org/wiki/Novikov_self-consistency_pri...

  • More generally, imposing "self-consistency" on a closed cycle of interactions is just a matter of picking a fixed point. Such a fixed point will always exist if the underlying system is continuous - and continuity may always be assumed if the system be non-deterministic. (For example, a billiard ball enters a wormhole sending it to the past with probability 50%, or else it is knocked away by a billiard ball sent from the future (and does not enter the wormhole) with probability 50%. This system is self-consistent, but this is achieved by a "mixture" of two outcomes.)

    • Can the ball roll into wormhole, emerge in the past, hit its past self and stop, while its past self it knocked to roll into the wormhole, emerge in the past, hit its past self ...

      1 reply →

    • Could you elaborate on what you mean by fixed point? Fixed point of what? And what continuity and non-determinism are you referring to exactly?

      1 reply →

  • This paper (among some others that are referenced in this Wikipedia article) are also cited here and referenced.

Most 'time loops' in science fiction might better be described as time knots.

I think of https://en.wikipedia.org/wiki/Predestination_(film) which is much more complicated than the usual time travel scenario; presumably the protagonist leaves but doesn't really enter since the protagonist is their own mother and father (the matter that makes them up does enter since they eat and breathe the way everybody else does; thinking the story through I'd think if I was going to have such a miraculous and singular existence I'd rather be a fantastic creature of some kind [dragon?] as opposed to a relatively boring intersex person capable of both reproductive roles)

Also https://en.wikipedia.org/wiki/The_End_of_Eternity which tames the complexity of time travel by presupposing 'eternity' has a second time dimension, making large-scale engineering of history practical. 'Eternity' itself owes it's existence to a time loop which is ultimately broken by the protagonist.

> a NEARLY closed timelike curve, which arrives from the past, coils around itself one or more times - like a coil, indeed - allowing causal interaction between the different spires (i.e. one can interact with its future self/selves and with its past self/selves), and finally the last spire leaves toward the future.

The classic sci-fi story describing this is Heinlein's By His Bootstraps. Note, though, that even in this version, the causal interactions are fixed: the same person experiences the events multiple times from different viewpoints, but the events have to be the same each time. They can't change. In Heinlein's story, the main character tries to do something different at one of these interactions and finds that he can't.

>which by the way only exists because you traveled back in time in the first place

No, you keep going forward all the time, but on a dimension closed on itself.

That's the whole point.

  • If time is closed on itself, then by definition there can be no change from one "round" to another, you have to return to the exact world state you started in. Otherwise it wouldn't be closed. Just like a coil is not a closed shape even if its projection (a circle) is.

Isn't the cool part of this the assertion that the arrow of time flips at points of minimum and maximum entropy? In other words, it's two parallel timelines, not a continuous loop of entropic time. The article dedicates itself to proving this assertion with a bunch of math of which I understood maybe 10%.

I am not a physicist, etc so if I sound daft then that's why.

Your younger self doesn’t have to be a future state of your present self, you just have to induce it, eg, being your own father or grandfather. Your younger self doesn’t have to be in your future if you allow some overlap (father) or short gap (grandfather) on the circle.

This reminds me of comic strips where someone interacts with an adjacent frame.

futurama solved that: fry travelled back and became his grandfather's younger self, rather than his own younger self