You have to give up the idea of a universal clock: you only have a local one. In your example: since you erased the elapsed computation in other regions, you (

You have to give up the idea of a universal clock: you only have a local one. In your example: since you erased the elapsed computation in other regions, you (by putting the particles back) are literally rewinding every clock but yours. In whatever sense they were ever in 2000, they're back there, though you're in (your) 2018. Having your 2018 be their 2000 seems like a sensible definition of time travel. (And yes, you'd have spent a year time traveling, though in many senses you always are time traveling forward, and you just happened to spend a year going fast "backwards".) For the rest of your points, I agree that they're sensible objections to being able to do it, but they don't make the concept not mean anything, which was your original point. Why would the future remain? You destroyed it when you time traveled, and any changes in the future would require time traveling forward to see. (Forward is less erasure heavy, because you're just stalling instead of rewinding.) It's not just a similar experience, it's the literal definition of what that term means. > nor time, just changes in particle positions Uh... It's comments like this which makes me question your position. Changes in particle positions is time! Literally: time is just that thing a clock does, which is tick. If particles change position, that's (by definition) time.