When I was in high school I read the novel "Lightning", by Dean Koontz, a thriller / potboiler with time travel, but I remember even then thinking it was uncommonly smart, because it essentially went through the reasoning you describe above when the time-traveling protagonist explains why travel to the past is impossible (and it's a big twist when this protagonist reveals he is not from the future, but from the past). It still suffers from a variant of the problems you describe, because time travelers can still return to the past (albeit always later than they left).
Time travel can be made metaphysically coherent in the following way: Marty steps in the time machine in 2100, then the time machine replicates the universe of 1900 a few moments later, but Marty is still in 2100, it just happens that 2100 has been changed to look and feel like 1900, including replicating all people alive in 1900.
So, time machines are actually universe make-over machines.
How about this: Marty goes to a parallel universe which was identical to ours in every way until the moment Marty appeared in it. He appears in it 100 years before the time in which he left his original universe. Granted this isn't exactly time travel, but it is very similar, and avoids all the causality issues. This seems metaphysically possible, and David Lewis' personal time seems like the most reasonable description of what's happening.
I don't find these arguments especially convincing. It feels like you're assuming there are no major physical phenomena that we don't already know about ------ other than time travel. In the same way that Newtonian mechanics is a limiting case of special relativity which is a limiting case of general relativity, how can we be certain e.g. conservation of energy isn't such a limiting case?
With cause and effect, isn't the obvious, common-sense notion that effects cannot precede causes predicated on the assumption that time travel cannot happen? Would it be so obvious if we don't start with the assumption time travel cannot happen? Cause and effect, as you're obviously aware, is a notoriously hard thing to pin down. I don't find it particularly helpful here.
Plus, while I'm not a physicist, if my limited understanding of quantum mechanics is correct, precisely predicting how certain quantum systems will behave *requires* that you (mathematically, if not philosophically) allow for (some) particles to move backwards in time during the U process. Happy to be corrected here by someone more familiar.
Either way, I think a more powerful idea than cause and effect where time travel is concerned is that of self-consistency. Again with regards to quantum mechanics and quantum entanglement, I suspect ideas like Einstein's spooky action-at-a-distance, the many worlds hypothesis or the Copenhagen interpretation are somewhat misguided. I think what quantum entanglement tells us about the universe is only that the only thing it promises is that carefully made observations will always be *consistent* with one another. One should certainly be suspicious about situations which appear to violate C&E, but a situation that is not logically self-consistent is a non-starter.
Whether or not C&E can be used to make sense of time travel, there are spacetime topologies and histories (where 'history' here must include the present and future) that at least seem to be self-consistent.
While I agree that time travel as it's usually portrayed in SciFi is almost certainly impossible due to problems with causality, I think your metaphysical arguments are missing the point.
In particular, argument 2 is just a problem of definitions. We _know_ that time is relative and can flow differently for different objects. Furthermore, depending on your frame of reference two events can even switch their order, for one observer event A will happen before B, and for another event B before A. (Provided locations of events A and B are far enough from each other for them not to have any causal connection.)
So a priori there's nothing immediately impossible in one observer's past being another observer's future. In physics time travel is formalized as "closed timelike curve" (https://en.wikipedia.org/wiki/Closed_timelike_curve), avoiding any ambiguities of terms "past" and "future".
Thanks. But even in relativity, there wouldn't be the sort of relativity involved in time travel -- e.g., there's no reference frame in which an event on Earth in 1900 is later than an event on Earth in 2100. The 1900 event is in the absolute past of the 2100 event.
The cases where events have different time order in different reference frames are always events outside each other's light cones. Which normally prevents them from influencing each other ... unless you have a case of quantum entanglement. But then, quantum entanglement probably just shows that relativity is false.
I agree that special relativity doesn't allow for travelling back in time into the same spot, but general relativity does. I still think that time travel is impossible in practice, but not for the reasons that you are listing.
For instance you are writing about "absolute past", but the assumption that there is an absolute timeline completely depends on time travel being impossible, thus making argument 2 tautological.
Argument 3 depends on the definition of the sameness. Also the statement that an object can't be in two places at once is far from obvious. For example, it's clearly false in quantum mechanics.
Argument 4 relies on the global conservation of energy, which is not a thing. Energy is only conserved locally in a closed system. Also depending on how exactly a hypothetical time machine works, it might not lead to any problems with the conservation of energy.
Argument 5 is close to what I think is the real reason time travel is impossible, but it's still formulated from the assumption of an absolute timeline, making it tautological.
A better formalization of the causality argument would be mathematical. If we imagine that the world is a differential equation with some starting conditions (which is not far from the truth), then it will usually have a single solution. Adding an additional condition that a part of the solution at time t1 should exactly match a part of the solution at time t2 would turn the equation unsolvable. It's similar to a system of three equations with only two variables: it will usually have no solution because the number of constraints is higher than the number of degrees of freedom.
Thank you! You undid a great confusion in my mind with this last paragraph about entropy.
People like to say that entropy creates time, or is the source of time. I now understand the source of the confusion.
When I was in high school I read the novel "Lightning", by Dean Koontz, a thriller / potboiler with time travel, but I remember even then thinking it was uncommonly smart, because it essentially went through the reasoning you describe above when the time-traveling protagonist explains why travel to the past is impossible (and it's a big twist when this protagonist reveals he is not from the future, but from the past). It still suffers from a variant of the problems you describe, because time travelers can still return to the past (albeit always later than they left).
Sounds cool, but that last part is weird.
Well, it was infinitesimally later.
Time travel can be made metaphysically coherent in the following way: Marty steps in the time machine in 2100, then the time machine replicates the universe of 1900 a few moments later, but Marty is still in 2100, it just happens that 2100 has been changed to look and feel like 1900, including replicating all people alive in 1900.
So, time machines are actually universe make-over machines.
How about this: Marty goes to a parallel universe which was identical to ours in every way until the moment Marty appeared in it. He appears in it 100 years before the time in which he left his original universe. Granted this isn't exactly time travel, but it is very similar, and avoids all the causality issues. This seems metaphysically possible, and David Lewis' personal time seems like the most reasonable description of what's happening.
I don't find these arguments especially convincing. It feels like you're assuming there are no major physical phenomena that we don't already know about ------ other than time travel. In the same way that Newtonian mechanics is a limiting case of special relativity which is a limiting case of general relativity, how can we be certain e.g. conservation of energy isn't such a limiting case?
With cause and effect, isn't the obvious, common-sense notion that effects cannot precede causes predicated on the assumption that time travel cannot happen? Would it be so obvious if we don't start with the assumption time travel cannot happen? Cause and effect, as you're obviously aware, is a notoriously hard thing to pin down. I don't find it particularly helpful here.
Plus, while I'm not a physicist, if my limited understanding of quantum mechanics is correct, precisely predicting how certain quantum systems will behave *requires* that you (mathematically, if not philosophically) allow for (some) particles to move backwards in time during the U process. Happy to be corrected here by someone more familiar.
Either way, I think a more powerful idea than cause and effect where time travel is concerned is that of self-consistency. Again with regards to quantum mechanics and quantum entanglement, I suspect ideas like Einstein's spooky action-at-a-distance, the many worlds hypothesis or the Copenhagen interpretation are somewhat misguided. I think what quantum entanglement tells us about the universe is only that the only thing it promises is that carefully made observations will always be *consistent* with one another. One should certainly be suspicious about situations which appear to violate C&E, but a situation that is not logically self-consistent is a non-starter.
Whether or not C&E can be used to make sense of time travel, there are spacetime topologies and histories (where 'history' here must include the present and future) that at least seem to be self-consistent.
That being said, I'm not holding my breath.
While I agree that time travel as it's usually portrayed in SciFi is almost certainly impossible due to problems with causality, I think your metaphysical arguments are missing the point.
In particular, argument 2 is just a problem of definitions. We _know_ that time is relative and can flow differently for different objects. Furthermore, depending on your frame of reference two events can even switch their order, for one observer event A will happen before B, and for another event B before A. (Provided locations of events A and B are far enough from each other for them not to have any causal connection.)
So a priori there's nothing immediately impossible in one observer's past being another observer's future. In physics time travel is formalized as "closed timelike curve" (https://en.wikipedia.org/wiki/Closed_timelike_curve), avoiding any ambiguities of terms "past" and "future".
Thanks. But even in relativity, there wouldn't be the sort of relativity involved in time travel -- e.g., there's no reference frame in which an event on Earth in 1900 is later than an event on Earth in 2100. The 1900 event is in the absolute past of the 2100 event.
The cases where events have different time order in different reference frames are always events outside each other's light cones. Which normally prevents them from influencing each other ... unless you have a case of quantum entanglement. But then, quantum entanglement probably just shows that relativity is false.
I agree that special relativity doesn't allow for travelling back in time into the same spot, but general relativity does. I still think that time travel is impossible in practice, but not for the reasons that you are listing.
For instance you are writing about "absolute past", but the assumption that there is an absolute timeline completely depends on time travel being impossible, thus making argument 2 tautological.
Argument 3 depends on the definition of the sameness. Also the statement that an object can't be in two places at once is far from obvious. For example, it's clearly false in quantum mechanics.
Argument 4 relies on the global conservation of energy, which is not a thing. Energy is only conserved locally in a closed system. Also depending on how exactly a hypothetical time machine works, it might not lead to any problems with the conservation of energy.
Argument 5 is close to what I think is the real reason time travel is impossible, but it's still formulated from the assumption of an absolute timeline, making it tautological.
A better formalization of the causality argument would be mathematical. If we imagine that the world is a differential equation with some starting conditions (which is not far from the truth), then it will usually have a single solution. Adding an additional condition that a part of the solution at time t1 should exactly match a part of the solution at time t2 would turn the equation unsolvable. It's similar to a system of three equations with only two variables: it will usually have no solution because the number of constraints is higher than the number of degrees of freedom.
I hope your substack venture succeeds, and I have subscribed. But I urge you to keep a free option that eventually includes all the serious content.
Doesn’t cause and effect count as metaphysics also?
The free subscribers get to see the post tomorrow. They should also be able to see all the archives.
Yes, theories about causality are part of metaphysics.