Mitch, Hal, and Eugene keep coming dangerously close to a full
discussion of SF level time travel, paradoxes and all.
Is there anyone on this list who has not at some time or other
personally tried to work out a theory of time travel that
would resolve, or escape from, the paradoxes?
One idea I had turns out (so I like to think) isomorphic to
what David Deutsch suggests! First, DD wrote an article in
Scientific American five or six years back that purported to
describe how time travel would really work. You can't imagine
my anticipation of reading that, only to be rather disappointed.
It seemed that DD was saying that --- well, you know how anything
can happen in the multiverse? E.g., your keyboard could suddenly
turn into a brown rabbit, because there is a small amplitude in
QM for just such a transformation? --- he was saying, or so it
seemed to to me, that the appearance of a time traveler from
the future was exactly such a miraculous transformation. How
But I always have to read Deutsch twice to appreciate him. In
"The Fabric of Reality" he devotes an entire chapter to time
travel! And he's right on the money (as we Americans say),
merely fleshing out his earlier SA article.
But here was my idea from many years ago: suppose (to make
the paradox simpler) we have a Contrarian. This person will
always say "Yes" if he is supposed to say "No" and vice-verse.
The Contrarian receives information from the future that he
is about to say "Yes", so naturally he says "No". Well, then
after he writes down his answer somewhere, and some time travel
apparatus gets hold of it, it's now the "No" answer that flies
backwards in time. This time, the past Contrarian is presented
with the "He said No" information, and so says "Yes". Then,
later, when **his** information is written down and then used
by the time travel apparatus, the cycle repeats.
As you have already probably guessed, this picture involves
a sort of meta-time. (You should properly have gagged at a
number of the "this time's" and "later's" that I used in
that paragraph.) Now then, we mark the "nth" instance of the
information going backwards in time, and then the "n+1th"
instance and so on, where n and n+1 are elements of a non-
well-ordered set (so that there is no starting element).
But it gives rise to an ensemble-type picture: you look
at these two paths and see them both happening infinitely
many times, but identically each time, which really amounts
to them just happening once each (under a certain interpretation).
The more elaborate paradoxes are handled similarly. You kill
your grandfather, and then aren't born in the next "run", but
then history is different, someone else gets the time machine,
then kills his grandfather (or sends the 1937 Electrical
Engineering Handbook back to 1837), and a different history
happens, and then a different history happens, and then...
some large but finite number of times, until you are born
once more, and the entire cycle repeats.
But so far as I can make out, this is what David Deutsch is
in effect saying. If you "go back in time" you are going to
a different branch of the multiverse, that's all. (The
multiverse is a continuum, so it would lack my finite recursion
property that I used for illustration above, but that isn't
> ...you've got a loop of Final Cause (the Antipode of the First
> Cause) keeping reaching back, and restructuring the timeline.
> Iteratively. Not only is this bootstrap, this is autofeedback.
> The capabilities are nudged backwards with each increment. The
> initial kinetics dynamics is limited, but I have no idea how
> limited. Even if you're getting back "only" information, you
> might be very surprised what you might be getting back from
> scant few hours into the future.
It's not completely clear to me what he's saying, but the idea
I get out of this is as follows. Suppose that you are climbing
a hill at a certain angle theta. Then information from the
future arrives which enables you to climb at theta+iota. But
then since you are now on *that* slope, so to speak, you are
going to be receiving information from a more advanced future.
The effect of that information will be to now place you on a
theta+iota+iota slope, and so on, as long as you keep your machine
receiving from the tangential future. (The tangent space, for
those who know differential geometry, is a good metaphor here.)
For the metaverse, this would have the following implication.
Intelligent life (i.e. life that is capable of devising and
profiting from the use of time travel devices) almost always
goes off on an indescribable tangent (to badly mix metaphors).
That is, a Singularity is achieved in a way that we have not
so far discussed very much.
Anyway, so much for paradoxes: they can be dealt with in the
manner that I have suggested, namely by an appeal to the
multiverse which (I think!) is indistinguishable from the
meta-time idea above. (My reference to the 1937 Electrical
Handbook is taken from the marvelous old time travel short
story "Other Tracks", which has inspired almost all my time
This archive was generated by hypermail 2b30 : Mon May 28 2001 - 10:00:06 MDT