About those time machines that you can "make"...

I have seen my fair share of pop science articles about time travel and the like. And it often comes up that in THEORY you can build a time machine of sorts. Not something useful and ful like a Dr Who Tardis, but a time machine.

These “doable” time machines seem to fall more along these lines. Yeah, you build and you can go back in time to the day it was turned on, but no further or the like.

So, you build one in 2525. At some point in the future, you can go back to 2525. Or the time machine has say a 25 year “capacity”. In 2250 you can go back to 2525. In 2575 you can go back to 2550.

Isn’t causality STILL a problem with these?

Oh, yes.

There are at least four ways that a time machine might cope with causality.
1/ It might refuse to work. The Chronological Protection Conjecture suggests that time machines will collapse, or explode, or simply fail to work if we ever try to build one,

2/ You may be able to go back in time, but you can only perform certain actions, consistent with history, when you get there. Hence no Grandfather Paradox. This is Novikov Consistency.

3/ If you go back you can rewrite history, but you will create a new timeline with different events in it. You will be forced to live in that timeline from that point onwards (unless you still have access to the time machine).
4/ In a many-histories, many worlds universe, you can go back in time, but you will live from that point onwards in a new branch of the infinitely branching universe, even if you don’t change history. Of course, there would be an infinitesimal chance that you could find your original worldline, but most of your branching instances would not. Good luck time-travelling in an infinitely branching universe, as you would be lost from day one (what ever day one means in this context).

I’ve been thinking about this and I think there is a way you can have something that might resemble a time machine without any breaches to causality.

Here are the rules of the time machine.

  1. You can only affect events that are inside your light cone at the time you enter the time machine. So you can’t go back in time to a point before you entered the time machine but you might be able to go to the future affect something and then come back and live the rest of your life now or somewhere in the future. You can also bring back whatever neat artifacts/people from the future that you want.
  2. It is possible to take free will actions, that prevent what you saw and did in the future from taking place. In which case the future is re-written and your actions are voided.

The basis of this time machine, is to imagine

  1. a omniscient computer that can accurately predict the future based on current conditions.
  2. an omnipotent device that act in a time delayed fashion to project an image of you to take certain actions at a future date, and which can manufacture anything that was predicted to exist in the future into the present time when you leave the time machine.

Neither of these two devices violate any known laws of physics or causality, and yet put together they are equivalent to the time machine previously described. That is not to say that the time machine must be implemented using these omnipotent and omniscient components, just that if they don’t disobey the laws of causality than your time machine won’t either.

Too late to edit:
In this machine the personal perspective time of the traveler is the important thing in terms of what actually happens in reality, with the latest action taking precedence. So for example you could enter the time machine tomorrow. Come out in 2100 AD, pick up a hitchhiker, take him back to 2050, and kill his grandfather. your hitchhiker would still be around, and doing stuff in 2050, but he wouldn’t actually be born because his grandfather died. But all of this is OK because it all happens after 2016 when you first entered the time machine, and the hitchhiker is just a hypothetical construct.

There is a more fundamental problem with “time machines” that is rarely addressed, and that is that pre-existing mass or energy that is sent backwards in time accumulates at the terminus point in spacetime. Not withstanding of where the additional mass came from, there is no limit to the amount of mass that can be sent backward up to the light cone of the origin of the time machine itself. If the time machine is some kind of open portal, there is no reason mass could not be send backward in a continuous stream, and because the mass would then be part of the past it could be recursively streamed in an infinite loop, increasing the mass at the origin without limit. This is, for obvious reasons, worse than any supposed paradoxes assuming singular causality and probably one of the most fundamental arguments against traversable closed timelike curves (though it does not prevent the existence of classes of Gödel metrics that are globally hyperbolic). Conservation of mass-energy is, of course, an assumption, but one that is so basic that finding a global violation that would allow for practical time travel would not only seriously compromise general relativity but essentially all of thermodynamics and the fundamental physics that rely on it.


That’s right. The inhabitants of the far distant future might not be too keen on living through the big rip/big crunch/heat death of the universe, and they might want to come back in time to the earlier, more clement phases of the universe.

Of course they could do this an arbitrary number of times, making the conservation of mass problem even more acute.

And I thought “snow birds” were irritating.

While such an accumulation of mass would doubtless have many interesting effects, it would not constitute a violation of conservation of mass nor of energy. The Law of Conservation of Energy does not actually say that energy cannot increase. It just says that if the energy inside a box changes, that amount of energy must have passed through the walls of the box. Assuming that the time machine functions as a portal or wormhole of some sort, the energy would be entering the box through the portal.

That could be a fun idea for a sci-fi story. Scientists discover that all black holes are the result of alien civilizations inventing the time machine and creating an infinite mass loop that collapses into a singularity. Since FTL travel is impossible, civilizations inevitably turn to time travel as a work around and it always ends in a singularity disaster. That’s the answer to the Fermi paradox.

Causality is only a problem in the sense that we have never yet observed an effect which precedes it’s cause (emphasis on the word YET). Based on what we’ve seen so far, we have developed a law which says causes must precede effects. But we have no proof that this law is somehow necessary for the universe to operate. It may be true that bidirectional-causality universes are possible, and no one has proven that our universe isn’t one.

It has been suggested that a bidirectional-causality universe might cope with the grandfather paradox by the fact that you can never actually succeed at killing your grandfather. You can try, but something will always get in your way. The gun will jam. The bullet will miss. You’ll misidentify who your grandfather was and kill the wrong person. There’s always a chance you would fail, even if the chance is tiny. In such a universe, amazingly unlikely things would happen to protect the timeline.

That reminds me of a book I read called The Great Time Machine Hoax. They build a really powerful computer and holodeck then asked the computer to simulate the past. The computer does such a good job at creating the simulation that it convinces itself that they actually had traveled back in time, and refuses to take them home because home no longer exists.

This is so-called Novikov consistency, which I mentioned in my previous post.
I think this consistency is quite popular with physicists who think that a limited form of time-travel is possible, but this concept seems to deny the possibility of free-will.

Some people are fine with that, it seems; the whole concept of free-will is a tricky one, in any case.

I love this. All this babble about something that no one really understands - time. Even the legendary **Stranger **can only talk about conditions that would “compromise general relativity,” etc. Sounds like some boys in the basement talking about the future. Seems to me that the fact that all the big minds who deal with these things still are no closer to the concept than they ever were means that the notion of time is the big baffler in all this. As Walter Von der Wogelweide said, “Yes - we can easily tell you what time isn’t, but we can’t tell you what it is.”

Time isn’t the difficulty. English is the difficulty. We can describe time just fine, in the appropriate language.

Ok, except that we don’t live inside mathematical formulas, but we live in a real world and we communicate and interact in non-mathematical language. And in that context, we can’t say what time is. We don’t even know what it is.